Date post: | 13-Feb-2017 |
Category: |
Documents |
Upload: | truongtram |
View: | 224 times |
Download: | 0 times |
Research Collection
Doctoral Thesis
Hydrogenation of aliphatic nitriles over nickel catalysts modifiedby formaldehyde
Author(s): Novi, Roc
Publication Date: 2004
Permanent Link: https://doi.org/10.3929/ethz-a-004877494
Rights / License: In Copyright - Non-Commercial Use Permitted
This page was generated automatically upon download from the ETH Zurich Research Collection. For moreinformation please consult the Terms of use.
ETH Library
DISS. ETHNO. 15708
Hydrogénation of Aliphatic Nitriles over
Nickel Catalysts Modified by Formaldehyde
A dissertation submitted to the
SWISS FEDERAL INSTITUTE OF TECHNOLOGY ZURICH
for the degree of
Doctor of natural sciences
presented by
ROC NOVI
Dipl. Chem. ETH
born on 15. October 1977
citizen of Vignogn GR
accepted on the recommendation of
Prof. Dr. P. Rys, examiner
Prof. Dr. M. Morbidelli, co-examiner
Zürich 2004
Page h
Acknowledgements
I wish to express my sincere gratitude to...
...Prof. Dr. P. Rys, who gave me the chance to work on this project. I am
very grateful to him for many discussions and his continuous support.
...Prof. Dr. M. Morbidelli for accepting co-examination of this doctoral
thesis.
...Dr. F. Rössler for the possibility to perform the experiments in
Kaiseraugst, the scientific support and the discussions. His confidence in
my work was very supporting and encouraging.
...Dr. A. Rössler and B. Sägesser for the numerous discussions, tips and
advices.
...my colleagues of the Rys group, especially Dr. F. Antognoli,
Dr. E. Dedeoglu, Dr. A. J. Klaus, Dr. M. Mösche, Dr. A. Rössler,
D. Schoch and Dr. P. Skrabal.
...the members of the hydrogénation team in Kaiseraugst, H. Bruder,
B. Close, A. Dodane, H. Lehmann, T. Müller, A. Saaler, B. Sägesser,
R. Santillo, P. Schmidt, G. Weisser and C. Zürcher for the good team
work.
...the members of the analytical groups in Kaiseraugst and Zürich
H. Kleissner, Dr. G Schiefer and Dr. P. Skrabal for the coaching and
measurements they performed.
...M. Bäbler, L. Brändli, M. Bucher, S. Diezi, M. Günther, J. Fischesser,
P. Müller and S. Sasso for the discussions and practical advice.
Page m
.my friends, D. Günther, I. Netzer, M. A. Plaz and J. W. Solèr for their
friendship and moral support apart from the work.
.my parents, my grandparents as well as my brother and sister for their
support during all these years of education.
.the company F. Hoffmann-La Roche for the financial support of this
project as well as the possibility to perform the experiments in its
laboratory in Kaiseraugst.
Page lv
Table of contents
1 Abstract 1
2 Zusammenfassung 3
3 Introduction 5
3.1 Overview 5
3.2 Aim and scope of this thesis 7
3.3 List of abbreviations and symbols 8
3.4 List of substances 9
4 Theoretical section 13
4.1 Hydrogénation of nitriles 13
4.1.1 General aspects 13
4.1.2 Hydrogénation of nitriles to amines 13
4.1.3 Hydrogénation of nitriles to aldehydes 15
4.1.4 Hydrogénation of nitriles to hydrocarbons 16
4.1.5 Hydrogénation and cyclisation 16
4.2 Mechanistic considerations of nitrile hydrogénation 17
4.2.1 Historic development 17
4.2.2 Models for the formation of side products 23
4.2.3 Influence of reaction parameters on nitrile hydrogénation 25
4.3 Reversible reactions of amines 27
4.4 Raney nickel 30
4.4.1 Preparation methods 30
4.4.2 Industrial applications 31
4.4.3 Variation of the properties of Raney nickel 32
4.4.4 Adsorption of nitriles and their hydrogénation
intermediates 33
4.4.5 Inhibition/poisoning of the catalyst 35
4.4.6 Acid sites 36
5 Hydrogénation of butyronitrile 39
5.1 Reaction system 39
Page v
5.2 Thermodynamic aspects 45
5.3 Aspects of mass and heat transport 45
5.4 Influence of reaction parameters on nitrile hydrogénation 47
5.4.1 Influence of the overall pressure 48
5.4.2 Influence of the temperature 50
5.4.3 Influence of the ratio of catalyst to substrate 54
5.4.4 Recycling of the catalyst 57
5.4.5 Influence of additives 59
5.5 Influence of washing/modification procedures 60
5.5.1 Influence of washing procedures with different solvents 60
5.5.2 Influence of the modification by formaldehyde 63
5.6 Reversibility of the hydrogénation steps 68
5.6.1 Disproportionation of butylamine 68
5.6.2 Influence of reaction parameters on the disproportionation
of butylamine 70
5.6.3 Dibutylimine as starting material 72
5.7 Discussion 73
5.7.1 The bifunctional catalytic hydrogénation and its
reversibility 73
5.7.2 Influence of various reaction parameters on the selectivity ...74
6 Modification of nickel catalysts by formaldehyde 77
6.1 General remarks 77
6.2 Influence of the treatment with different solvents on the
properties of the catalyst 79
6.2.1 Reduction potential 79
6.2.2 Adsorption of an indicator 80
6.3 Modification of Raney nickel by various formaldehyde
concentrations 81
6.3.1 Analysis of the modifying solution 81
6.3.2 Properties of the modified catalysts 84
6.4 Modification of various amounts of Raney nickel at constant
modification strength 85
6.4.1 Analysis of the modifying solution 85
6.4.2 Properties of the modified catalysts 87
Page vi
6.5 Modification of nickel-on-carrier 89
6.6 Discussion 89
7 Effect of formaldehyde modified nickel catalysts on other chemical systems ...91
7.1 Hydrogénation of crotonaldehyde 91
7.1.1 General remarks 91
7.1.2 Test for a possible gas-liquid transfer limitation for
hydrogen 93
7.1.3 Influence of the formaldehyde modification of Raney
nickel on the hydrogénation of crotonaldehyde 94
7.1.4 Influence of the formaldehyde modification of
nickel-on-carrier on the hydrogénation of crotonaldehyde ...96
7.2 Hydrogénation of l-bromo-4-nitrobenzene 97
7.2.1 General aspects 97
7.2.2 Test for a possible gas-liquid transfer limitation for
hydrogen 98
7.2.3 Influence of the modification on selectivity and reaction
rates 99
7.3 Hydrogénation of levodione 101
7.3.1 General remarks 101
7.3.2 Test for a possible gas-liquid transfer limitation for
hydrogen 101
7.3.3 Influence of the modification on selectivity and
hydrogénation rate 102
7.4 Discussion 103
8 Conclusions and outlook 105
9 Experimental 107
9.1 Apparatus 107
9.1.1 Description of the 500 ml steel hydrogenator 107
9.1.2 200 ml glass hydrogenator 109
9.1.3 100 ml low pressure hydrogénation apparatus 109
9.1.4 Lab Shaker 110
9.1.5 Modification and washing apparatus 110
9.1.6 Gas Chromatograph 110
9.2 Methods ill
Page vu
9.2.1 Hydrogénation of butyronitrile Ill
9.2.2 Reversibility experiments in the 35 ml screening autoclave 111
9.2.3 Hydrogénation of crotonaldehyde 112
9.2.4 Hydrogénation of l-bromo-4-nitrobenzene 112
9.2.5 Hydrogénation of levodione 113
9.2.6 Description of the sampling procedure 113
9.2.7 Neutralisation with water 114
9.2.8 Neutralisation with methanol 114
9.2.9 Neutralisation with tetrahydrofurane 114
9.2.10 Modification with formaldehyde 114
9.3 Analytics 115
9.3.1 Determination of butyronitrile, butylamine, dibutylamine
and dibutylimine with a GC method using an internal
standard 115
9.3.2 Determination of crotonaldehyde, crotylalkohol, butanal
and butanol with a GC method using an internal standard...
116
9.3.3 Determination of l-bromo-4-nitrobenzene, l-bromo-4-
aminobenzene and aniline 118
9.3.4 Determination of levodione and actinol 119
9.3.5 Methanol determination in aqueous medium with a
headspace GC method using an external standard 120
9.3.6 Formaldehyde determination in aqueous medium with an
HPLC method using an external standard 121
9.3.7 Synthesis of dibutylimine as a standard for GC
measurements 122
9.4 Identification of by-products 123
9.4.1 N-Butylbutanamide 123
9.4.2 N,N-Dibutylbutyramidine 123
9.4.3 1,1-Diethoxybutane 124
9.5 Characterisation of the catalyst 124
9.5.1 The reduction potential 124
9.5.2 Adsorption of an indicator 125
9.5.3 Dissolution in acidic medium 125
9.6 Chemicals 125
Page vin
9.7 Calculation of selectivity and reaction rates 127
9.8 Error analysis 128
9.8.1 Precision of a gas chromatographic analysis 128
9.8.2 Precision of a sample analysed with gas chromatography ...
129
9.8.3 Precision of the selectivity and the reaction rates 130
10 Literature 131
11 Appendix 137
11.1 Curriculum vitae 137
11.2Conference contributions 138
Pagex
Chapter A
Abstract
In the present work the liquid phase hydrogénation of aliphatic nitriles as
well as an industrially applicable modification procedure for hydrogénation
catalysts [1] producing higher yields of primary amines were investigated.
In a first part, by using Raney nickel as catalyst and butyronitrile as a
model substance the influence of the gas-liquid hydrogen transfer limitation
and several reaction parameters, such as the temperature, the hydrogen
pressure, the ratio of catalyst to substrate, the recyclability of the catalyst
and various additives on the hydrogénation selectivity were investigated.
The influence of these parameters is discussed with a semi-quantitative
macro-kinetic model presented within this thesis. Furthermore, the
reversibility of the reaction steps that characterise the hydrogénation system
was investigated with an intermediate product as starting material.
A new, economically interesting and easily applicable method to rise the
selectivity towards primary amines is the modification of nickel catalysts by
formaldehyde. Several parameters of this modification process were
investigated, because the desired higher selectivity is also accompanied by
an undesired loss of activity during the hydrogénation. In particular the
amount of formaldehyde used to treat the catalyst and the amount of catalyst
at a constant formaldehyde concentration were explored. The nickel
leaching during the modification as well as during the hydrogénation was
shown to be also an economically relevant factor, if a modified catalyst is
used in an industrial process.
The effect of the modified catalysts on other chemical systems was
screened by employing the following hydrogénation systems: the
hydrogénation of an oc,ß-unsaturated aldehyde, the hydrogénation of a
halogenated nitroarene and the enantioselective hydrogénation of a cyclic
Page 1
dione. The modification by formaldehyde was not beneficial in the tested
cases as the sélectivités were not enhanced. On the contrary, a decrease in
activity was observed with the modified catalysts.
Page 2
Chapter £â
Zusammenfassung
Im Rahmen der vorliegenden Arbeit wurden die Hydrierung von
aliphatischen Nitrilen in flüssiger Phase sowie ein industriell anwendbarer
Prozess zur Modifizierung von Hydrierkatalysatoren [1], welcher höhere
Ausbeuten an primären Aminen liefert, untersucht.
Anhand der Hydrierung von Butyronitril mit Raney Nickel wurde neben
dem Gas-Flüssig-Stofftransport für Wasserstoff auch der Einfluss der
Temperatur, des Wasserstoffdrucks, des Verhältnisses von Katalysator zu
Substrat, der Rezyklierbarkeit des Katalysators und von Zusatzstoffen auf
die Selektivität untersucht und mit Hilfe eines semi-quantitativen makro¬
kinetischen Modells diskutiert. Ferner wurde die Reversibilität der
Reaktionsstufen des Hydriersystems untersucht, indem ein Zwischen¬
produkt als Substrat verwendet wurde.
Die Hydrierung mit Katalysatoren, die mit Formaldehyd modifiziert
wurden, ist eine neue, ökonomisch interessante Möglichkeit zur Steigerung
der Selektivität bezüglich primären Aminen. Zur Implementierung der
modifizierten Katalysatoren sind jedoch mehrere Parameter zu untersuchen,
da die erwünschte Selektivitätssteigerung von einem unerwünschten
Aktivitätsverlust begleitet wird. Insbesondere die Formaldehyd¬
konzentration, aber auch das Katalysator-Substrat-Verhältnis sowie deren
Auswirkungen auf eine gegebene Nitrilhydrierung wurden untersucht. Es
wurde festgestellt, dass für eine industrielle Anwendung der Katalysatoren
das Herauslösen von Nickel in die Modifizierlösung wie auch in die
Hydrierlösung ein wichtiger Kosten bestimmender Faktor ist.
Die Einwirkung der modifizierten Katalysatoren auf andere chemische
Systeme wurde ebenfalls evaluiert. Hierbei wurden die modifizierten
Page 3
Katalysatoren auch bei der Hydrierung von oc,ß-ungesättigten Aldehyden,
halogenierten Nitroaromaten und Dialdehyden getestet. Es stellte sich
heraus, dass die Formaldehyd-Modifizierung für diese Systeme keinen
Vorteil in Form einer Selektivitätssteigerung bringt: Es konnte lediglich ein
Aktivitätsverlust beobachtet werden.
Page 4
Chapter +J
Introduction
3.1 Overview
The catalytic hydrogénation of nitriles leads to a mixture of primary,
secondary and tertiary amines, amides and alcohols. The economic
efficiency of this catalytic reaction to the primary amines is influenced
mainly by the by-products mentioned above. Further research is, therefore,
necessary to determine an optimal selective hydrogénation for each nitrile.
Formation of the by-products can be minimized by using a suitable catalyst,
by modification of the catalyst, or by additives such as ammonia or
alcohols. The most important catalyst for this system is Raney nickel and its
modifications, but the exact mechanism, especially the intrinsic selectivity
determining step, is presently unknown. The aim of this work is, therefore,
to obtain a deeper understanding of the mechanism and to achieve an
optimization of the process.
The selectivity obtained during nitrile hydrogénation is determined by a
strong interaction of two catalytic functions, the acid and the hydrogénation
sites. This is shown in the following reaction scheme (Figure 3-1). The ratio
of and the distance between the acid and the hydrogénation sites determine
the ratio of the reaction constants kH4/kC5, and therefore the formation of
by-products (secondary and tertiary amines). It can be assumed that the
reaction pathway via kH3, kH4 is kinetically favoured towards the reaction
pathway via kH1, kH2. Thus, a minimal number of acid catalytic sites is
favoured.
Page 5
H2/Ni H,/Nir. n R CH
© Àkm
© À+ H + H
„
V*s
V
©
r ni-
H2/Ni ©
R CH
NH - R—CH,
^H2
H,/Ni
NH,
^H3
+ RCH2NH2
h:?
^H4
R CH NH2
©
H2N CH2 R
©
R CH NH?
HN CH2 R
A
©
R CH
-NH,
©
+ H
- R—CH,
©
-NH-,
HN CH2 R
H,/Ni R CH2
H2N CH2-©
©
H
RCH2NHCH2R
tertiary amine
Fig. 3-1 : Bifunctional mechanism for the hydrogénation of nitriles.
Page 6
3.2 Aim and scope of this thesis
The aim of this work is to investigate the mechanism presented in
Chapter 3.1 and to ultimately confirm or reject its feasibility. Therefore,
investigations on the following parameters have been performed:
catalyst
catalyst modification
hydrogen pressure
additives
temperature
solvent
In addition, a recently patented modification of hydrogénation catalysts
by formaldehyde, which leads to better selectivities towards the primary
amines if aromatic nitriles are hydrogenated, is tested for aliphatic nitriles.
The completely unknown mechanism of this modification has also to be
investigated.
Page 7
3.3 List of abbreviations and symbols
Table 3-1 : List of abbreviations and symbols.
actinol (4R,6R)-4-hydroxy-2,2,6-trimethylcyclohi
BA butylamine
BN butyronitrile
BI butylimine
cat catalyst
Damide dibutylamide, JV-butylbutanamide
DBA dibutylamine
DBD dibutylamidine
DBDH+ protonated dibutylamidine
DBI dibutylimine, dibutylazomethine
EDS l,2-bis(2-hydroxyethylthio)ethane
ESTD external standard
FAMEC8 n-caprylic acid methyl ester
FID flame ionisation detector
ISTD internal standard
levodione (6R)-2,2,6-trimethylcyclohexa-l,4-dione
MeOH methanol
RaNi Raney nickel
Tamidine tributylamidine, iV,iV-dibutylburyramidine
TBA tributylamine
TFA trifluoro acetic acid
THF tetrahydrofuran
C concentration
exp experiment
Page 8
Table 3-1 : List of abbreviations and symbols.
G gas phase
kx rate constant of the reaction x
L liquid phase
P(H2) hydrogen pressure
P(NH3) ammonia pressure
r distance
ro initial reaction rate
G standard deviation
rel. G relative standard deviation
S solid phase
T temperature
3.4 List of substances
Table 3-2 : List of substances.
/ V^x/ \
-NH2 4-aminoazobenzene
CH30
4-amino-5-cyano-2-methoxy-
pyrimidine
pynitrile
Page 9
Table 3-2 : List of substances.
-NH, aniline
HS O
l,2-bis(2-hydroxyethylthio)ethaneSH
Br-
Br-
-NH9 l-bromo-4-aminobenzene
-NO, l-bromo-4-nitrobenzene
butanol
NH
N
butyraldehyde
butylamine
butylimine
butyronitrile
n-caprylic acid methyl ester
Page 10
Table 3-2 : List of substances.
crotonaldehyde
crotylalkohol
dibutylamide
NH;
N
H
NH
dibutylamidine
protonated dibutylamidine
dibutylamine
iV,iV-dibutylbutyramidine
tributylamidine
Tamidine
Page 11
Table 3-2 : List of substances.
(4R,6R)-4-hydroxy-2,2,6-trimethyl-
cyclohexanone
actinol
HO' R
ypyridine
tributylamine
(6R)-2,2,6-trimethylcyclohexa-l,4-
dione
levodione
Page 12
Chapter ^
Theoretical section
4.1 Hydrogénation of nitriles
4.1.1 General aspects
The reduction of organic substances with molecular hydrogen in the
presence of solid contact substances is based on heterogeneous catalysis.
Because the substances do not react with hydrogen under standard
conditions, the reaction occurs exclusively on the contact surface. Different
processes, which depend on the adsorption of reaction partners on the
catalyst, and which are not understood in detail, determine the selectivity of
such reactions. Consequently, the result of the reduction often depends on
the catalyst and the chosen conditions.
4.1.2 Hydrogénation of nitriles to amines
The hydrogénation of nitriles is one of the main methods used in industrial
chemistry for the synthesis of primary amines. The product is always a
mixture of primary, secondary and tertiary amines. The selectivity of this
catalytic reaction depends on several factors: The nature of the catalyst
(metal and support), the addition of ammonia, the temperature and the
solvent used [2-4]. This was impressively demonstrated in the case of
butyronitrile {Figure 4-1) [5].
To control the selectivity towards the primary amine, the following
methods are proposed in the literature [6] :
Page 13
Ni / diatomacous earth (50%),
NH3, CH3OH, 125°C
Rh/C(5%),
NH3,H2O,75-110oC
Pt/C(5%),
NH3, H20, 125°C
Fig. 4-1 : Selectivities of butyronitrile hydrogénation using different catalysts,
conditions and solvents [5].
• Hydrogénation in acid solution: The produced amines are converted
into salts and thereby deactivated for a further reaction to higher
substituted amines.
• Hydrogénation under acylating conditions: In reactions with acetic
anhydride or methyl formate the primary amines are converted into
amides, which subsequently can be hydrolysed.
• Hydrogénation in presence of ammonia: Reactions that produce
secondary and tertiary amines by the cleavage of ammonia are inhibited
in the presence of ammonia.
In industrial nitrile hydrogénation the addition of ammonia is the method of
choice, which has no great disadvantages and produces considerable higher
selectivities towards the primary amine. A new, industrially applicable
method to increase the selectivity to primary amines is to modify the
catalyst by formaldehyde (see Chapter 4.4.3).
Page 14
4.1.3 Hydrogénation of nitriles to aldehydes
The hydrogénation of nitriles produces aldimines as intermediates.
Therefore, in principle it is possible to synthesize aldehydes via the partial
hydrogénation of nitriles and the subsequent hydrolysis of the produced
aldimines. This conversion is only successful with y- and 8-hydroxynitriles,
because the cyclisation of the corresponding aldimines to the saturated
2-aminotetrahydrofuranes and -pyranes is faster than the further
hydrogénation (Figure 4-2) [7, 8].
ÇN HÇ^NH H2N HO
CH CH(ÇH2)n (ÇH2)n
Pd,H2
-OH -*--*-
(CH2)n Ö H?0 (CH2)n' O
-OH -*-
R R HR HR
n = 2-3
Fig. 4-2: Hydrogénation and cyclisation of a hydroxynitrile [6, 8].
Based on these fundamentals, in sugar chemistry a process for chain
prolongation was developed. Hydrogen cyanide is added and the
intermediate is partially hydrogenated and subsequently hydrolysed.
The conditions for an intramolecular stabilisation do not exist in most
aliphatic and aromatic aldimines. Nevertheless, the nitrile hydrogénation
can be stopped after the uptake of one mole equivalent hydrogen if suitable
basic compounds are added and the secondary imine is produced. In
general, the more stable the formed secondary imines are, the higher is the
conversion.
Aldehydes can be produced from the secondary imines according to two
processes:
• Hydrolysis: Short heating with diluted mineral acids in alcohol or acetic
acid. The azomethines are split into aldehydes and the salts of the
released amines.
Page 15
• Displacement: Aldehydes which produce more stable secondary imines
displace the other aldehydes.
Another possibility, the direct hydrogénation of nitriles to aldehydes in
acidic medium, was investigated by Möltgen and Tinapp [9]. High
selectivities were obtained with Raney nickel as catalyst in acid solutions.
4.1.4 Hydrogénation of nitriles to hydrocarbons
With a mixed catalyst of nickel- and copper(II)-oxide (3:2) on silicic acid,
nitriles can be converted to hydrocarbons, e.g. p-aminobenzonitrile to
p-toluidine (80%) or cinnamic acid nitrile to propylbenzol (90%) [4, 6]. For
this conversion, the catalyst is prehydrogenated at 300°C and subsequently
a mixture of nitrile and hydrogen is added at similar temperatures. This
transformation is also possible with a molybdenum sulphide catalyst [6].
4.1.5 Hydrogénation and cyclisation
Cyclisations occur in catalytic hydrogénations if a functional group is
produced that can react with a second functional group located in suitable
distance. One of the best known examples is the reduction of y-hydroxyl-
acids to y-lactones. Schiffbases or primary amines in molecules containing a
second interacting group, such as ketones, esters, acids, amides, olefins and
various heterocyclic rings can be hydrogenated into cyclic amines [2, 3]. A
second nitrogen group can also lead to cyclic amines via the hydrogenolysis
of ammonia. Halogen atoms in the y- or 8-position favour cyclisations to
pyrrolines or piperidines.
Cyclisations are not limited to catalytic hydrogénation, but can also occur
during organic reductions. Because the favourable conditions for
cyclisations are not identical to the conditions for hydrogénation, the
cyclisation sometimes occurs when the reaction mixture is recovered. In
some cases, the cyclisation can be either suppressed or enforced in presence
of large quantities of ammonia or acylating solvent, respectively [2, 3, 6].
Page 16
4.2 Mechanistic considerations of nitrile hydrogénation
4.2.1 Historic development
In the literature results are generally discussed according to a mechanism
proposed by Braun et al. [10] in 1923, based on a competition between
heterogeneous hydrogénations and homogeneous condensations
(Figure 4-3, Figure 4-4).
H2 H2
R- EN :NH
RR NH2
Fig. 4-3: Hydrogénation of the nitrile, producing the intermediate imine which is
further hydrogenated to the primary amine.
With his experiments Braun et al. [10] could exclude the reaction pathway
to the secondary amine via the aldehyde intermediate formed by the reaction
of aldimine with water.
H9:NH-
RCH2NH2
NH?H2
NH,
R R
route A
NH,
,^\,
H2
N^ R R
route B
Fig. 4-4: Mechanism proposed by Braun et al. [10] for the reaction towards the
secondary amine.
Page 17
Two possible intermediates (Figure 4-4) were proposed, the 1-amino-
amine (route A) and the secondary imine (route B). After the reaction of the
nitrile with one mole equivalent of hydrogen, the aldimine is produced,
which can then react with hydrogen to produce a primary amine. As soon as
both the aldimine and the primary amine are present, they react to form the
Schiff'base by condensation to the 1-aminoamine and splitting of ammonia.
The SchiffbasQ can then be hydrogenated to the secondary amine (route B).
Alternatively, the amine and the aldimine can react to form the
1-aminoamine, which can be hydrogenolysed to the secondary amine (route
A).
In 1967, Greenfield [11] presented a similar scheme for the formation of
the tertiary amine (Figure 4-5). The aldimine reacts with the secondary
amine to produce the 1-aminotrialkylamine, which is hydrogenolysed to the
tertiary amine (route A). Greenfield also proposed the alternative reaction
path via the intermediate of the enamine (route B).
y-NH(CH2CH2R)2 R
=NH «*~
Fig. 4-5: Mechanism for the tertiary amine formation proposed by Greenfield [11].
In 1986, VolfandPasek [12] summarised the hydrogénation in a scheme
and pointed out that there exist two types of reactions: Typical hydrogéna¬
tions (A, B, E and H) and acid-base catalysed condensations (C, D, F and G,
Figure 4-6).
Page 18
R NH2
B H,
NH2
RCHoNH,-=N :5=t D/^ ^=^
NH2
R N
NH,
-NH,
/^KR ^NH r
R N f<nR N
HN(CH2R)2
,R "2
R Ni/
-R N
Fig. 4-6: Scheme for the hydrogénation process proposed by VolfandPasek [12].
In 1992, Dallons et al. [13] postulated a mechanism for the formation of
by-products in which a semi-hydrogenated intermediate adsorbed on the
catalyst reacts with a primary or a secondary amine. The resulting 1-amino-
alkylamine or 1-aminodialkylamine reacts further to the secondary imine or
the tertiary enamine (Figure 4-7).
R1
H
H9EN
^R
^N
NHR2CH2R1« R1
M]
NH2
NR2
-R1
M]
Fig. 4-7: Intermediates, chemisorbed on a metal of the catalyst, that lead to secondary
(R2=H) or tertiary amines (R2=CH2R1) in a mechanism proposed by Dallons
et al. [13].
Non-hydrogenating active sites (acid sites) are responsible for the
adsorption of the amine on the catalyst. Obviously, the support has a great
Page 19
influence on the selectivity. The adsorption of primary and secondary
amines as well as ammonia is enabled by the acid sites. Dallons et al.
proposed that acid sites on the support give a high selectivity in favour of
the primary amines because these amines are adsorbed on the sites that are
not neighbouring the hydrogénation sites. The formation of by-products is
therefore suppressed.
A fundamentally different mechanism was formulated by Verhaak et al.
[14-16] {Figure 4-8). The hydrogénation of acetonitrile in the gas phase was
investigated using various acid nickel catalysts. The acidity of the catalysts
was successfully decreased by modifying the reaction temperature and by
the addition of potassium as a promoter to achieve a higher selectivity
towards the primary amine. The quantity and strengths of the acid sites were
determined by temperature programmed desorption (TPD) of ammonia, and
a linear correlation was found between the selectivity and the quantity of the
acid sites.
In this mechanism, the imine produced during the hydrogénation can
either be further hydrogenated to a primary amine or desorb from the
catalyst (Figure 4-8). If the imine readsorbs on acid sites, the acid-catalysed
gas phase metallic function
H,
= R- :NH
^ R CH2—NH2
migration(gas phase)
-R CH2—N CH2—R*
acidic function
H+
R CH—NH2
^ RCH9—NH,
i2—i\in2
NH9
R CH2—NH2-CH—R
NH,
^ R CH2—N^=CH—R
Fig. 4-8: Mechanism proposed by Verhaak et al. [14-16] in which side reactions to
secondary and tertiary amines are catalysed by acid sites on the catalyst.
Page 20
side reactions occur (Figure 4-9) forming azomethine and enamine. These
substances then migrate back to the hydrogénation sites through the gas
phase and are hydrogenated to secondary or tertiary amines.
NH
©H
© NH,NH2
II
NH2
NH2
R'
NH2
©NH,
R'
-H
NH2
NH
R'
©
NH3
N
©-H -NH,
Vc/R
Fig. 4-9: Condensation reactions on the acid sites of the catalyst: I Chemisorbed
primary imine on the acid sites and its resonance structures, II reactions of the
primary imine to produce the secondary imine or the 1-aminoamine [16].
Huang andSachtler [17-22] presented a mechanism based on deuterium
exchange experiments. Acetonitrile was hydrogenated with D2 in the
presence of a ruthenium catalyst to produce CHD2CN and CH3CD2NH2
(Eq. 4.1). Deuterated acetonitrile (CD3CN) was hydrogenated with H2 to
produce CD3CH2ND2 and CDH2CN (Eq. 4.2).
2 CH3CN + 2 D2 CHD2CN + CH3CD2NH2 Eq. 4.1
Page 21
2 CD3CN + 2 H2 CDH2CN + CD3CH2ND2 Eq. 4.2
Based on these experiments, Huang and Sachtler formulated a
mechanism without an imine intermediate but instead with the formation of
intermediates that are chemisorbed by double bonds on the catalyst
(Figure 4-10). According to Huang et al. [18] and Rode et al. [23], acid sites
CN
+
CH3
CD2
N
Ru
CH3
„CN CD, CN
|+
CH3
CD2
1 ^CH3
CH2, NH
\/Ru
CH
Ru
NH2
Fig. 4-10: Mechanism proposed by Huang and Sachtler that explains the H2/D2 distri¬
bution observed in the experiments [22].
on the catalyst surface have no influence on the selectivity towards the
primary amine. This conclusion is based on two ideas [18]:
Secondary and tertiary amines are also formed on catalysts on neutral
supports.
Acid sites on a catalyst are neutralised in strongly basic media.
Coq et al. [24] postulated a mechanism (Figure 4-11) which includes
the side reactions and does not exclude Huang and Sachtler's [22]
mechanism. The mechanism of Verhaak et al. [14] is also not excluded. The
side reactions that lead to the secondary and tertiary amines can take place
on acid sites or on hydrogénation sites of the catalyst. A mechanism for the
condensations on the hydrogénation sites was postulated (Figure 4-11), in
which a primary amine reacts with the chemisorbed carbene E, or
alternatively the chemisorbed intermediates D and E condense.
Page 22
CH,
CH^=N
H
R r. N -r. n
* *
A
^" B
N
**
C2^.2
2H
CH2 NH2
* *
D
R NH2
**
E
^
Fig. 4-11 : Mechanism for the hydrogénation proposed by Coq et al. [24].
4.2.2 Models for the formation of side products
Nucleophilic attack on the imine carbon: Addition of water to the double
bond of the imine, forming an aldehyde via the aminole intermediate
(hydrolysis of the imine). This aldehyde then condenses with amines to
secondary imines that are hydrogenated to the corresponding amines
(Figure 4-12) [25, 26].
R CH^NHH,0
OH
R CH NH2
-NH, /O
R Cv
H,N^ ^R
H,
R NHCH^N,
OH
VH2° R C H
HN.
Fig. 4-12: Formation of by-products: Side reaction with water, forming the secondary
amine [26].
Page 23
Nucleophilic attack on the cyano carbon: Nucleophiles (water, alcohols
and amines) can attack the nitrile group (Figure 4-13).
H
I ,0H20 | /s
***"
[M] N •<*" [M]-—NH2—(5'
^C OH R
R
H
HOR' |[M] N^=C R <
*"[M] N
^C OR'
R
H
I /NR'NH,R' | //
<»
[M] N «» [M]~--NH2—C.
NC NHR' R
R
Fig. 4-13: Formation of by-products: Nucleophilic attack on the cyano carbon [26].
Insertion reactions: Insertion of the nitrile on chemisorbed
intermediates and by-products (Figure 4-14).
Electron transfer and C-C coupling reaction: One electron transfers
from the nitrile to the metallic catalyst, forming an iminoradical that can
dimerize (Figure 4-15).
Obviously, the reactions to hydrocarbons (at high temperatures) and the
cyclisation reactions (Chapter 4.1.4 and Chapter 4.1.5) can also diminish
the yields of primary amines.
Page 24
H NCR' /R[M] N CH2R < *"- [M] N^=C
R'
NCR' /[M] R <
"[M] C.
N CH2RH
N R
/ NCR' /[M]^C „
*•[M]^C X
R N=
NCR'
R
R'
[M]—R « [M]—N=C^R
Fig. 4-14: Formation of by-products: Insertion reactions [26].
R
C^=N [M]n
[M]n N^^C R m" [M]*""1'—N^=C*
<" [M]n N^=C
R R
Fig. 4-15: Formation of by-products: One electron transfer and C-C coupling reaction
[26].
4.2.3 Influence of reaction parameters on nitrile hydrogénation
Influence of ammonia
The effect of ammonia on the selectivity was first investigated by Braun et
al. [10]. All authors describe a higher selectivity towards the primary amine
if ammonia is added to the reaction mixture, but the reason of this effect is
unknown. Another unexplained effect is that the reaction rates are higher if
ammonia is added up to a certain ammonia concentration and are then again
lowered [27]. Several explanations were put forward:
Page 25
• Ammonia influences the equilibrium between the amine, the imine and
the azomethine (Figure 4-16) [28].
CH2 CH -NH3 ^. .CH
R NH2 R-^ ^nh **R N^ R
Fig. 4-16: Equilibrium of amine, imine, azomethine and ammonia [28].
• Ammonia reacts with the primary imine to produce the 1 -aminoamine,
which is hydrogenolysed to form the primary amine {Figure 4-17) [29].
<ZH NH3R ^NH -•
NH2
H, CH,- »- / ^
~* R NH2
NH,
Fig. 4-17: Addition of ammonia to the double bond of the imine and hydrogenolysis of
the 1-aminoamine [29].
Ammonia poisons the acid sites of the catalyst, inhibiting the acid-
catalysed side reactions [16].
Ammonia modifies the electronic properties of the catalyst, preventing
the unwanted side reactions.
A similar positive effect on the selectivity can be observed if alkali
hydroxides are added to the reaction mixture [30, 31].
Influence of the solvent
Besson et al. [32] investigated the effect of various solvents on the
selectivity and the activity by comparing the polarity of the solvent with the
selectivity towards the primary amine. It was found that the more polar the
solvent, the higher the selectivity. In addition, the activity in different
alcohols was investigated, and it was found that the reaction rate increased
Page 26
as the number of C-atoms in the alcohol increased. The solubility of
hydrogen in the solvent also influences the hydrogénation reaction.
Influence of the temperature
Generally, the selectivity decreases as the temperature increases [26].
Degischer [27] found that within the range 60-140°C the selectivity
increased linearly with increasing temperature. At temperatures above
160°C, the selectivity decreased with increasing temperature. This change
may be due to different activation energies at higher temperatures (kinetic
control) or a shift towards equilibrium at higher temperatures
(thermodynamic control).
Influence of the hydrogen pressure
A higher hydrogen pressure causes a higher reaction rate. Furthermore, a
higher pressure, in general, leads also to a higher selectivity. However, if
Raney nickel is employed, a higher hydrogen pressure surprisingly yields a
lower selectivity towards the primary amine [26].
Influence of the catalyst
The main influence on selectivity and activity is based on the metal catalyst
and its support. Often similar selectivities and activities are observed in the
liquid and the gas phase hydrogénation.
4.3 Reversible reactions of amines
The synthesis of secondary and tertiary amines starting from primary
amines was described by Nicodemus and Schmidt [33] in 1930. Ethylamine
and butylamine reacted at temperatures of about 220°C to the
corresponding secondary amines using a catalyst produced by coating
cobalt carbonate onto pumice stone. Selectivities of 76-78% towards the
secondary amine were observed at conversions of 65-70%. 1936 Herold and
Smykal [34] reported on the preparation of primary amines from secondary
Page 27
and tertiary amines and ammonia. At temperatures of 300-450°C, using an
excess of ammonia and catalysts such as alumina gel, activated carbon,
aluminium oxide, reduced nickel catalysts, yields up to 66% were achieved.
A commercially interesting application is the production of hexamethylene-
diamine starting from azepane, that was patented by Reppe and Bauer [35]
(Figure 4-18). In the presence of hydrogen, ammonia and a nickel or cobalt
Fig. 4-18: Production of hexamethylenediamine starting from azepane (I) and the
corresponding secondary amine (II).
catalyst azepane is converted to hexamethylenediamine "in good yields" in
the liquid phase at temperatures of 140-220°C. This reaction is of interest
because azepane is produced as a by-product during the hydrogénation of
adiponitrile. Also, the production of hexamethylenediamine from the
corresponding dimer and ammonia is an interesting reaction since the dimer
is another by-product of the adiponitrile hydrogénation (Figure 4-18).
The reverse reaction, the cyclisation of hexamethylenediamine, was
also investigated [36], and azepane was obtained at conversions of 84% in
the gas phase at temperatures of 350-380°C using chromium and vanadium
oxide catalysts. Nowadays, the production of secondary amines starting
with primary amines is one of the standard methods in the laboratory
[37-40].
Page 28
In 1993/94 Verhaak et al. [14-16] proposed a mechanism for the
hydrogénation of amines. Their attention focussed on the acid sites on the
catalyst and their role in the production of higher substituted amines as
by-products (vide Chapter 4.2.1, Figure 4-8 and Figure 4-9).
In another publication, the disproportionation of the propylamine in the
gas phase using a continuous flow reactor, hydrogen and hydrogénation
catalysts was investigated [41]. The production rate of dipropylamine
formation decreased with increasing hydrogen pressure. If the reaction was
run without hydrogen, the conversion to the secondary amine was decreased
and dipropylimine as well as propylimine were obtained as main products
(Figure 4-19). The reaction rates depend on the acid sites on the catalyst.
Fig. 4-19: Production of dipropylamine B (in the presence of hydrogen), or dipropyl¬
imine C and propylimine D (in the absence of hydrogen) [41].
A mechanism was proposed in which the disproportionation is divided
into four reaction steps (Figure 4-20): a dehydrogenation followed by an
acid catalysed condensation and finally the hydrogenolysis of the amino-
amine produced.
This reaction mechanism is supported by the fact that the number of the
acid sites determined by temperature programmed desorption of ammonia
Page 29
correlates with the conversion of propylamine in the disproportionation
experiments (Figure 4-8).
-NH3
Fig. 4-20: Reaction sequence for the disproportionation of primary amines proposed by
Verhaaketal. [41].
4.4 Raney nickel
4.4.1 Preparation methods
Since the work of Sabatier [42] nickel is known as a good hydrogénation
catalyst. To enlarge the reaction surface, the metal was dispersed on
inorganic supports. Another way to increase the activity was discovered by
Raney [43] in 1925. His patent describes a process to remove Si from a
NiSi-alloy with alkaline solutions. NiAl intermetallic components showed a
higher activity than those of Si [44]. The investigation of Raney nickel is
interesting due to its complex skeletal structure and its wide range of
application in organic synthesis. Raney nickel is often used in industry, e.g.,
in the catalytic hydrogénation of adiponitrile to hexamethylenediamine, or
in the hydrogénation of benzene to cyclohexane.
Page 30
Raney nickel is prepared by leaching the Al in a NiAl-alloy with a
sodium or potassium hydroxide solution in accordance with Eq. 4.3 [45,
46]. If the alkaline hydroxide is not used in large excess (20-30% NaOH or
30-40% KOH), the aluminate formed is deposited as bayerite on the catalyst
(Eq. 4.4).
2 Al + 2 OH" + 2 H20 ^=^ 2 A102" + 3 H2 Eq. 4.3
2 A102" + 4 H20 =5=*= A1203 x 3H20 + 20H_ Eq. 4.4
There are three methods that can be used to manufacture Raney nickel. In
the first method, the nickel particles are slowly added to an alkaline
solution. In the second method, the solution is slowly added to the alloy in a
neutral suspension. In both cases, it is important that the reaction is
controlled. If aluminium has to be removed quantitatively, the nickel
particles must be added slowly to an alkaline solution. When no further
hydrogen is evolved, the reaction mixture is heated in concentrated alkali
solution.
The concentration of the leach decreases with conversion, so that the
suspension must be decanted several times, and the lye replaced. The fresh
Raney nickel is stored in a 1 M sodium hydroxide solution. To minimize
variations in the properties of the catalyst, all samples should be taken from
the same batch.
4.4.2 Industrial applications
Raney nickel is a heterogeneous catalyst often used, with many applications
in hydrogénation reactions [47-49], especially in the hydrogénation of
nitriles to primary amines which are used in polymeristion reactions. This is
due to the good selectivity and the cheap price of the metal compared to
other metals used in hydrogénation processes. Another field of application
is the hydrogénation of aromatic nitro compounds [50], of C-C double
bonds and even as a substitute for the Lindlar catalyst (Table 4-1).
Page 31
Table 4-1 : Examples for the applications ofRaney nickel in industrial processes [47-49].
reaction substrate product application
functional group
hydrogénation 2,4-dinitrotoluene 2,4-diaminotoluene polyurethanes
nitro
hydrogénation 1,5,9-cyclododeca- cyclododecane polyesters from
diene triene nylon-6,12
hydrogénation C10-C13-3-ketoacid Cio"Ci3-3- pharma products
ketone hydroxyacid
hydrogénation 2-ethylhexanal 2-ethylhexanol plasticiser
aldehyde
hydrogénation stearonitrile stearylamine plasticiser
nitrile
hydrogénation adiponitrile hexamethylenedi - nylon-6,6
dinitrile amine
hydrogénation 1,4-butynediol 1,4-butanediol THF
alkyne
hydrogénation benzene cyclohexene polyamides
aromatic
hydrogénation phenol cyclohexanol polyamides
aromatic
aminolysis 1,6-hexanediol hexamethylenedi - nylon-6,6
alcohol amine
alkylation dodecylamine dimethyldodecyl- surfactants
amine amine
4.4.3 Variation of the properties ofRaney nickel
Various methods were tested to obtain more selective catalysts for
different processes, mainly in the hydrogénation of nitriles and of aromatic
nitro compounds. There, the main parameters are the metal composition
Page 32
(ratio of nickel/aluminium) [50-52], the strength of the basic treatment, the
doping with other metals [53-55], the process parameters during the
production process (e.g. quenching the parent alloy in cold water [56-58])
and the modifying additives such as lithium hydroxide [31], morpholine
[59], copper acetate [60], formamidine salts [61] or vanadium salts [62]. A
recently published patent of Degischer and Rössler [1] is presenting the
advantages of a catalyst modified by formaldehyde in the hydrogénation of
nitriles. A yield of 96.4% was obtained if 'pynitrile' (4-amino-5-cyano-2-
methoxypyrimidine) was hydrogenated using a commercially obtainable
Raney nickel. The selectivity was increased to 99.6% if the catalyst was
treated with a 1% formaldehyde solution prior to the hydrogénation
reaction. Similar effects on the selectivity were found if instead of
formaldehyde carbon monoxide (98.8% primary amine) or acetaldehyde
(97.3% primary amine) were used to modify the catalyst [63]. Further
experiments with benzonitrile as a model substance are presented in
Figure 4-21. The disadvantage of such a modification is the loss of activity,
which is leading to a higher catalyst load in industrial processes [64].
In addition, experiments with Raney cobalt and nickel-on-carrier
catalysts were patented. Again, the modification by formaldehyde has a
positive effect on the selectivity towards the primary amine. The selectivity
for the primary amine increased from 96.8% to 98.1% if a nickel-on-carrier
was modified using formaldehyde, and from 96.8% to 99.1% if a Raney
cobalt catalyst was modified.
4.4.4 Adsorption of nitriles and their hydrogénation intermediates
Recently, great efforts have been made to describe the adsorption of
hydrogen as well as of other substances involved in the catalytic
hydrogénation of nitriles. Semi-empirical studies, molecular modellings,
studies on metalorganic substances as well as spectroscopic data (high
resolution electron energy loss vibrational spectroscopy: HREELS) were
reported [26, 65-69]. De Bellefon and Fouilloux [26] summarised the
chemisorbed species in a scheme as shown in Figure 4-22.
Page 33
ww
CO
E
0
g
Ero
>^N£Z
0
100
no modification
1%CH20 modified
2.5% CH20 modified
5% ChLO modified
120 140 160 180 200
reaction time / [min]
Fig. 4-21 : Hydrogénation of benzonitrile using a differently modified catalyst. Reaction
conditions: 100 ml benzonitrile, 30 g ammonia, 2 1 methanol, 5.7 g Raney
nickel, 100°C, 4 MPa and 1200 rpm [1, 64].
Because not only the knowledge of the interaction between nitriles and
their hydrogénation intermediates with the catalyst is important to
understand the hydrogénation, Blyholder and Neff [70] investigated the
adsorption properties of solvents, such as methanol, ethanol, diethyl ether
and water using a nickel-on-siliciumoxide catalyst. The adsorbed species
were observed with infrared spectroscopy. It was found that methanol reacts
at a temperature of 20°C and produces chemisorbed CO on the catalyst
surface. Ethanol reacts in the same way, so that in addition to the infrared
band of CO also bands of CH3 and CH2 were detected. Ni-CH3,
Ni-CH2-CH3 and Ni-0-CH2-CH3 are supposed to be chemisorbed at the
surface. Neither water nor diethyl ether chemisorbed or reacted with the
surface.
From these experiments Blyholder and Neff'[70] concluded that carbon
monoxide is the only species that is chemisorbed on nickel surfaces. The
Page 34
RCN RCN* RCN* + H RCN* + 2H RCN* + 3H RCN* + 4H RCN + 4H
R
C
R H R R
CH2 11 HCs
R
CH2
R
CH2
y y N
*1 ul N\
N XNH NH NH2
/ [M] [M] [M] [M] [M] [M]
' end-on nitrile metaloimrne amido end-on imine amino amine
R—CH=^NH R -CH2 NH2
free nitrile
[M]
side-on nitrile
\
[M]
side-on inline
free amine
\
R\C^NH R\ /NHl
H
R f ^NH2
[M] [M] [M]
lminoacyl aminocarbene aminoalkyl
Fig. 4-22: Catalytic hydrogénation of nitriles: Intermediates adsorbed on a metal centre
of the catalyst [M] [26].
reaction pathway of alcohols leading to CO produces aldehyde or ketone as
intermediates. Initially, the alcohol is dehydrogenated before the C-C and
the C-H bonds are broken and CO and other fragments are produced.
Diethyl ether does not react with nickel surfaces as it can not dehydrogenate
a hydroxyl group. The dehydrogenation of alcohols on Raney nickel was
also reported by Besson et al. [25].
4.4.5 Inhibition/poisoning of the catalyst
Every process, chemical or physical, that reduces the activity of a catalyst
can be classified as deactivation. In most cases, especially in complex
reactions or when using complex catalysts, a change in activity is
accompanied by changes in selectivity. A change in selectivity can be
achieved using poisons or inhibitors. The following processes are important
[71,72]:
Irreversible adsorption of an inhibitor at the surface or reaction of an
inhibitor with the catalyst surface.
Page 35
Competitive reversible adsorption of a poison on the surface.
Poison induced restructuring of the surface.
• Physical or chemical blocking of the pores.
With the example of a hydrocracking reaction, Penchev et al. [73]
demonstrated that the selectivity can be changed using different poisons
(thiophene poisons the metallic sites, pyridine the acid sites).
In the case of Raney nickel acetonitrile desorbs at temperatures above
75°C, while more strongly adsorbed nitrile fragments or molecules desorb
at temperatures above 180°C. The decomposition of acetonitrile forms two
carbon compounds that deactivate the catalyst. McCarty and Wise [74] and
Kock et al. [75] identified the adsorbed species as oc-carbon and nickel
carbide. These compounds can be hydrogenated at temperatures above
200°C, thereby restoring the catalyst's activity. Thus, the deactivation of the
catalyst during the hydrogénation of acetonitrile can be prevented by
applying high hydrogen pressures [24, 76].
4.4.6 Acid sites
The influence of acid sites on the selectivity of the hydrogénation of nitriles
is still disputed. Verhaak et al. [14-16] described the influence of acid sites
present during the hydrogénation of nitriles on the selectivity and the cata¬
lyst activity using nickel on a silicon and magnesium support. These acid
sites are formed during the synthesis of the catalyst and are ascribed to the
support. Their amount and strength were determined by the temperature
programmed desorption (TPD) of ammonia.
Increasing the number of acid sites leads to higher reaction rates and a
decreased selectivity. These observations were confirmed by Cabello et al.
[28] and Dung et al. [77].
Other authors [12, 17-22] did not confirm the influence of acid sites and
point out that the selectivity is influenced only by the metal used and by the
reaction conditions.
Aluminiumoxide is one of the most commonly used support materials in
heterogeneous catalysis. Examples are Pt-Re/Al203 (reforming) and
Co-Mo/Al203 (dehydrosulfonation). This support is thermally stable and
Page 36
allows an appropriate distribution of catalytically active compounds. How¬
ever, as aluminiumoxide is not an inert support, many reactions are cataly¬
sed by it, e.g. double bond migrations, E/Z isomerisations of olefins and
H/D exchange in hydrocarbons.
Knözinger [78] determined the acid sites via the infrared adsorption of
CO and measuring the stretching vibrations of C-0 and O-H bonds, which
depend on the bond strength (Figure 4-23). CO is also chemisorbed by
Lewis acid sites via G-bond and 7t-rebond. Knözinger determined the
oxidation state and the coordination number of cations on the surface and
the relative bond strengths of metal-CO bonds. In addition, acid site
sequences were determined for hydroxyl groups on the surfaces.
M—O—H— CO
Fig. 4-23: Chemisorption of CO on surface hydroxyl groups [78].
Page 37
Page 38
Chapter *J7
Hydrogénation of butyronitrile
5.1 Reaction system
The hydrogénation of butyronitrile leads to butylamine as main product if
Raney nickel is used as hydrogénation catalyst. A scheme of the reaction is
given in Figure 5-1. The hydrogénation experiments were made in a 500 ml
steel hydrogenator (see Chapter 9.1.1) according to the procedure described
in Chapter 9.2.1. The reversibility experiments presented in Chapter 5.6
were made in the 500 ml steel hydrogenator or in 35 ml screening
autoclaves (see Chapter 9.1.4) according to the procedure described in
Chapter 9.2.2. Two different catalysts were used, Raney nickel B 113 Z,
batch 20018989 from Degussa-Hillls AG and nickel-on-carrier Ni 1404 P,
lot H-99 form Engelhard.
Always when dibutylimine is mentioned, this means, that this is the
amount of dibutylimine measured by gas chromatographic analysis or
reaction rates extrapolated from gas chromatographic measured values.
However, one must keep in mind that it is possible, that other substances as
1-amino amines, amidines or imines react in the injector and the column of
the gas Chromatograph (at temperatures of 250°C) and then appear as
dibutylimine in the gas chromatographic spectrum.
Next to dibutylamine, tributylamine was observed as by-product.
Dibutylamide and the tertiary amidine were also formed in the reaction
mixture (Figure 5-2).
The product distribution of the hydrogénation is expressed in [mass-%],
the reaction rates in [mmol/s] or in [mmol/s*kg catalyst]. The indicated
pressure ascribes to the overall pressure, used during the reaction.
Page 39
CS
13 Q
C3
13 Q
+
4= »
Ö
J3
-&\43 P5
l-J-l ^t-
+ ^
<n
oN
43
-d
u
Ö
+
öo
43 P5
K
43
Fig. 5-1 : Hydrogénation of butyronitrile (BN) to the desired primary amine (BA) and
the secondary amine (DBA) over butylimine (not observed by GC) and
dibutylimine (DBI) as intermediate.
Page 40
tributylamine
TBA
NH
N-butylbutanamide, dibutylamideDamide
N,N-dibutylbutyramidineTamidine
Fig. 5-2: By-products of the hydrogénation of butyronitrile, observed by GC.
In order to follow through a well structured discussion of the influence
of various reaction parameters on the rate and the selectivity of the
hydrogénation of nitriles, first a simple semi-quantitative macro-kinetic
model is presented. The basis for the derivation of the kinetic equations is
the mechanism depicted in Figure 5-1. For the sake of simplicity it is
assumed that the rates of the individual hydrogénation steps are of the x-th
order with respect to the partial pressure of hydrogen. Furthermore, based
on the experimental data it can be assumed, that for the concentrations of
the intermediates BI, DBDH+ and DBD the steady-state approximation of
Bodenstein [79] can be applied.
The differential selectivity ratio (d[BA]/d[DBA]) is given by (Eq. 5.3):
d[BA]=
dt
d[DBA]
dt
k2P?H2)[BI]
k5P(H2)[DBI]
d[BA]_
k2 [BI]
d[DBA] k5[DBI]
Eq. 5.1
Eq. 5.2
Eq. 5.3
Page 41
The ratio [BI]/[DBI] can be evaluated from the following equations
(Eq. 5.5-Eq. 5.8):
^JP = k1[BN]PH2-[BI](k_1 + k2p^ + k3[BA][H+]) Eq.5.4
+k 3[DBDH+]
d[D^H ]= k3[BI][BA][H+] +k 4[DBI][H+]p(NH3) Eq. 5.5
-(k4 + k_3)[DBDH+]
As the protolytic side equilibrium between DBDH+ and DBD is assumed to
be much faster than all the other partial reaction steps, it can be neglected
for the formulation of the kinetic equations. From the equations (Eq. 5.4)
and (Eq. 5.5) the equations (Eq. 5.6) and (Eq. 5.7), respectively follow:
k^BNJp^ + k 3[DBDH+] -
ffiü
[BI] = — Eq. 5.6
k-l+k2P?H2) + k3([BA][H ])
k3[BI][BA][H+]+k4[DBI][H+]p(NH3)-d[D^H ]
[DBDH+] = —— - Eq. 5.7
K4 + K_3
If the steady-state approximation of Bodenstein [79] applies for the
intermediates BI, DBDH+ and DBD, i.e. if
d[BI]
dt«k1[BN]p^H2) + k_3[DBAD]
and
d[DBDH+] «k [BI][BA][H ]+k4[DBI][H ]p(NH
dt{ 3
Page 42
it follows:
_
k^tBNlp^j + MDBIHH Jp^ + kjkJBNJp^[B1J — Eq. 5.8
(k_3 + k4)(k_! + k2p^H2)) + k3k4[BA][H ]
and thus
[BI]_
1(k-3 + k4)k1[BN]p^)[5^] + k_3k_4[H+]p(NH3DBI] k4 (k 3
+ k4\ x +
1^J(k-1+k2P(H2))+ k3[BA][H]
Eq. 5.9
From the equations (Eq. 5.3) and (Eq. 5.9), the differential selectivity
(Eq. 5.10) follows:
d[BA] k2(k-3 + k4)k1[BN]P(H2)[5^ + k_3k_4[H+]p(NH3)Eq. j.lu
d[DBA] k4k5 As + k4^ x +
[-^J(k_1+k2p^H2))+ k3[BA][H ]
Equation (Eq. 5.10) reveals a very diversified and complex influence of the
selectivity on the various reaction parameters. A given reaction parameter,
such as e. g. the partial pressure of hydrogen, can either increase or decrease
the selectivity with respect to the desired butylamine, depending on the
relative rates of the individual reaction steps. This diversity shell be
exemplified by the following few cases:
Case I: k2p^2) » k_Y ; k_3 » k4 ; k_4[H+]p(NH3) » ki[BN]P*H2)i5^j
d[BA]_
k2k-3 k_4[R ]P(NH3d[DBA] k5 k_3k2p*H2) + k3k4[BA][H+]
Eq. 5.11
Case II: k2p*Ui) « k_Y ; k_3 » k4 ; k_4[H+]p(NH3) » ki[BN]P?H2)i5^7J
Page 43
d[BA]_
k2k_3 k_4[H ]p(NH3)e ^ ^
d[DBA] k5 k.gk^ + kgkJBAHH*]
Case III: k2p^2) » k_j ; k_3 « k4 ; k_3k_4[H+]p(NH3) » kik4[BN]P*H2)i5^7j
d[BA]_
k2k-3 k-4[H ]P(nh3)E 5 13
d[DBA] k4k5k2p^2) + k3[BA][H+]
Case IV: k2p*Ui) « k_Y ; k_3 « k4 ; k_3k_4[H+]p(NH3) » kik4[BN]P*H2)i5^7j
d[BA]_
k2k_3 k_4[H ]p(NHad[DBA] ^sk^+^CBA]^]
Case V: k2p^2) » k_Y ; k_3 « k4 ; k_4[H+]p(NH3) « ki[BN]P*H2)i5^j
d[BA] _k2k-3 kl[BN]p^)[DBli
Eq. 5.14
d[DBA] k5 k_3k2p^H2) + k3k4[BA][H+]Eq. 5.15
Case VI: k2p^2) « k_j ; k_3 » k4 ; k_4[H+]p(NH3) « ki[BN]P*H2)i5^j
d[BA]_
k2k_3 ki[BN]P?H2)[5Blid[DBA] k5 k^k^ + k^JBAHH^
Eq. 5.16
Page 44
Case VII: k2p*H2) » k_Y ; k_3 « k4 ; k_3k_4[H+]pNH3 « ^k^BN^2DBI
d[BA]_
k2 ki[BN]p^)ipiriEq 517
d[DBA] k5k2p^2) + k3[BA][H+]
Case VIII: k2p^2) « k_j ;k_3 « k4 ; k_3k_4[H+]p(NH3) « k4ki[BN]P*H2)i5^7j
d[BA]_
k2kl[BN]p^)î5ïïïiEq51g
d[DBA] k5k,+kJBA][H+]
5.2 Thermodynamic aspects
A profound investigation of the thermodynamic equilibrium of the
hydrogénation of butylamine within a temperature range of 25 to 150°C and
a hydrogen pressure range of 1 to 20 MPa partial pressure of hydrogen was
made by Chojecki [80] using HSC Chemistry software. The best
thermodynamic selectivity was calculated at high hydrogen pressures and
low temperatures (75% butylamine at 20 MPa hydrogen and 298 K). The
thermodynamic selectivity decreased dramatically at lower hydrogen
pressures, while temperature had only a small influence.
5.3 Aspects of mass and heat transport
The kinetics of multiphase reactions are not only influenced by the chemical
kinetic but also by the rate of mass and heat transfer. The investigated
system presented in this study is a three phase hydrogénation consisting of
gaseous hydrogen, liquid substrate and a solid catalyst. Mass transfer of
Page 45
hydrogen from the gaseous phase into the liquid phase as well as that of
hydrogen and the substrate (A) from the liquid phase to the surface of the
catalyst can significantly influence the kinetics of the individual reaction
steps, and thereby, also the selectivity of the overall hydrogénation process
(Figure 5-3).
catalyst S
Ca,s
Ch2,s
r
Fig. 5-3: Model of a three phase mass transport: Mass transport from the gaseous into
the liquid phase and from the liquid phase to the surface of the catalyst.
Cy: concentration of reagent i in phase j. i:H2 = hydrogen, iA = substrate A;
a,b,c: diffusional boundary layers.
Heat transfer from the surface of the catalyst to the liquid phase can also
influence the selectivity. This can occur if the reaction rates are much faster
than the heat transfer. In the case of a highly exothermic reaction such as the
hydrogénation of nitriles, the temperature on the surface of the catalyst is
higher than the measured temperature in the bulk liquid phase (Figure 5-4).
This, in turn, increases the local hydrogénation rate which than increases the
local temperature even further (run away). A different temperature effect on
C,, gas phase G
CH2,G
Page 46
Tu
liquid phase L
V
catalyst S
Fig. 5-4 : Temperature profile of an exothermic reaction: Heat transfer from the surface
of the catalyst to the liquid phase.
the rates of the competing reactions leads to a selectivity behavior which is
influenced by heat transfer.
5.4 Influence of reaction parameters on nitrile hydrogénation
The influence of the different reaction parameters on the initial differential
selectivity, the product distribution and the activity of the catalysts are
discussed in this Chapter. The product distribution (as yield butylamine after
full conversion) is plotted versus the initial differential selectivity
(calculated by dividing the initial production rate of butylamine d[BA]/dt by
the initial production rate of dibutylamine d[DBA]/dt) in Figure 5-5. It
seems quite evident that an interdependence of these parameters exists and
that with increasing values of the initial differential selectivity the yield of
butylamine approaches a maximum limiting value which is definitely lower
than 100%.
Page 47
92-
91 -
90-
,—,-
89-
CO -
CO
CD88-
F -
HC-*-~~~
Cl>"
c 86-
E -
CD 8b->s
3
.Q84-
Ti'
CD83-
>s-
82-
81 -
an-
parameter:stirrer speed
• mass catalystA temperatureV recycling
pressure
0 5 10 15 20 25 30 35 40 45 50
initial differential selectivity / [-]
Fig. 5-5: Initial differential selectivity versus product distribution (yield butylamine).
5.4.1 Influence of the overall pressure
The influence of the overall pressure on the product distribution, initial rates
and initial differential selectivity was investigated for 240 g BN, 3.75 g
RaNi, 100°C, and 1000 rpm in a range from 1 to 6 MPa. The results are
summarised in Figure 5-6, Figure 5-7 and Figure 5-8. The selectivity
towards the primary amine is decreasing with higher partial pressures of
hydrogen. Such a selectivity behavior is predicted by the approximate
equation (Eq. 5.10), mainly for the special cases I and III. The maximum
amount of the intermediate dibutylimine is almost independent of the
hydrogen pressure. The initial rates are rising using higher pressures.
The large scattering range of the measured initial reaction rates is due to
technical problems with the cascade regulation of the temperature. When
large amounts of hydrogen are pressed into the reactor, the temperature is
decreasing due to the cold gas. This is compensated by the temperature
regulation. If the reaction is then started, the reactor is heated from outside
as well as from inside by the exothermic reaction. The temperature
controlling cascade regulation is compensating this trend and starts to cool.
Page 48
au -
^
85-
-—B-^^B —
—m-ï£ 80-
CO
—— butylamine
—
CO
I75-• dibutylamine
tributylamine
o 70=.T- max. dibutylimine
-."
~V- T-
~vproductdistribut
O
Ol
O
.
I.I.I.-*- T-~T~ -
-
-
5-
0- 1 1 1 ' 1A A à. È. *
i '
2 3 4 5
overall pressure / [MPa]
Fig. 5-6: Influence of the overall pressure on the product distribution. Reaction
conditions: 240 g BN, 3.75 g RaNi, 100°C and 1000 rpm.
550-
500-
450-
400-
cd 350-h—»
cc
'Z. 300-
o
'5 250-
CD
£ 200-
CD
« 150-^g
100-
50-
0-
-d[BN]/dt / [mmol/(s*kg)]
• d[BA]/dt/[mmol/(s*kg)]* d[DBA]/dt/[E-1 mmol/(s*kg)]
d[DBI]/dt/[E-1 mmol/(s*kg)]
overall pressure / [MPa]
Fig. 5-7: Influence of the overall pressure on the initial reaction rates. Reaction
conditions: 240 g BN, 3.75 g RaNi, 100°C and 1000 rpm.
Page 49
28-
26-
24-
-22-1
I 20-
8 18-
CD
'£ 16-
CD
| 14-
1 12-
~
ID¬
S'
overall pressure / [MPa]
Fig. 5-8: Influence of the overall pressure on the initial differential selectivity.
Reaction conditions: 240 g BN, 3.75 g RaNi, 100°C and 1000 rpm.
However, due to a heat transfer delay the cooling is not fast enough. As a
consequence, a certain thermal runaway occurs. Thus, the difference
between the measured and the effective reaction temperature amounts in
some cases up to 10°C. This in turn has a massive effect on the initial rates.
5.4.2 Influence of the temperature
The influence of the temperature on the product distribution, the initial rates
and the initial differential selectivity was investigated in a range of 60-
120°C using standard conditions (240 g BN, 3.75 g RaNi, 1 MPa,
1000 rpm). The results are presented in Figure 5-9, Figure 5-10 and
Figure 5-11. The increase in selectivity towards butylamine can be
explained by the lower activation energy for the hydrogénation to dibutyl¬
amine. Activation energies were determined using a "quasi" Arrhenius plot
(Figure 5-12). The conditions of 240 g BN, 100°C, IMPa and 1000 rpm are
near the gas-liquid hydrogen transfer limitation so that the experiments (and
Page 50
95-1
90-
^85-
\ 80-
CD
Ê75-I
o 70'-4—'3
E 15 H
is iono
T3O
5-
0-
-5
butylamine
dibutylamine
max. dibutylimine
60 70 80 90 100 110 120
temperature / [°C]
Fig. 5-9 : Influence of the temperature on the product distribution. Reaction conditions:
240 g BN, 3.75 g RaNi, 1 MPa and 1000 rpm.
275-
250-
225-
200-co
CD
TO 175-
0 150-
1 125 H
ro 100-
^ 75-
50
25
0
-d[BN]/dt / [mmol/(s*kg)]
• d[BA]/dt / [mmol/(s*kg)]* d[DBA]/dt/[E-1 mmol/(s*kg)]
d[DBI]/dt / [E-1 mmol/(s*kg)]
60 70 80 90 100 110 120
temperature / [°C]
Fig. 5-10: Influence of the temperature on the initial rates of butyronitrile
hydrogénation. Reaction conditions: 240 g BN, 3.75 g RaNi, 1 MPa and 1000
rpm
Page 51
35 n
30-
-25-1
'>
'sM 20-CDCO
"cd
'E 15-
CD
=5 10-
"CD
5-
—1 1 1 1 1 1 1 1 1 1 1 1-
60 70 80 90 100 110
—I '
120
temperature / [°C]
Fig. 5-11: Influence of the temperature on the initial differential selectivity. Reaction
conditions: 240 g BN, 3.75 g RaNi, 1 MPa and 1000 rpm.
6.0-1
5.5-
5.0-
4.5-
4.0-
3.5-
3.3.O-
^2.5-
2.0-
1.5-
1.0-
0.5-
0.0'
0.0025
In (d[BN]/dt / [mmol/(s*kg)])• In (d[BA]/dt / [mmol/(s*kg)])* In (d[DBA]/dt / [E-1 mmol/(s*kg)])
In (d[DBI]/dt / [E-1 mmol/(s*kg)])
—i—
0.0026
1
0.0027
1 1 r
0.0028
1/T/[1/K]
—I—
0.0029 0.0030
1
0.0031
Fig. 5-12: "Quasi" Arrhenius plot obtained from the initial reaction rates r0 (see
Figure 5-10).
Page 52
Table 5-1 : Activation energies determined from the "quasi" Arrhenius plot by linear fit.
rates r0 temperature range activation energy error (±)
/ [see Figure 5-12] /[°C] / [kJ/mol] / [kJ/mol]
-d[BN]/dt 60-100 48.9 2.2
-d[BN]/dt 100-120 26.9 1.8
d[BA]/dt 60-100 50.7 1.8
d[BA]/dt 100-120 23.8 2.7
d[DBI]/dt 60-80 63.4 2.5
d[DBI]/dt 80-120 37.9 0.6
d[DBA]/dt 60-80 39.5 0.9
d[DBA]/dt 80-120 15.8 2.2
specially the reaction rates) using temperatures above 100°C are mainly
influenced by this limitation. The ln(-d[BN]/dt) versus 1/T data show a kink
in the linear correlation at a value of 0.00268 (100°C). The linear fit in the
range of 60 to 100°C gives an activation energy of 49 kJ/mol, above 100 °C
an activation energy of about 27 kJ/mol. Interesting data points were
obtained for the production rates of dibutylimine and dibutylamine. A sharp
break at the linear plot at 0.00283 (corresponding to 80°C) can be observed.
An explanation for these sharp breaks in the linear plots can be given either
by a diffusion limitation of the reagents to the corresponding hydrogénation
or acid sites, respectively, or by a change of the rate determining steps (see
Chapter 5.3). As the activation energy is approximately half as large at high
temperatures than at low temperatures the explanation of a diffusion
limitation is more probable. A summary of the obtained activation energies
is given in Table 5-1.
Page 53
5.4.3 Influence of the ratio of catalyst to substrate
The influence of the catalyst to substrate ratio on the product distribution,
the initial rates and the initial differential selectivity was investigated to
exclude a gas-liquid transport limitation for hydrogen. The following
reaction parameters were chosen: 240 g BN, 100°C, 1 MPa and 1000 rpm.
The catalyst amount was varied from 1.25 g to 20 g (0.5-8.3% catalyst). The
results are plotted in Figure 5-13, Figure 5-14 and Figure 5-15. The
T ! { ! { ! { ^
0 5 10 15 20
mass catalyst / [g]
Fig. 5-13 : Product distribution using different amounts of catalyst. Reaction conditions:
240 g BN, 100°C, 1 MPa and 1000 rpm.
selectivity towards the primary amine is rising using larger amounts of
catalyst. This may be due to catalyst deactivation (see Chapter 5.4.4). The
initial reaction rates are rising linearly with the catalyst amount until 5 g
catalyst. This is an indication that the hydrogen concentration in the liquid
phase is constant and no mass transfer limitation exists for hydrogen within
this range. Above a catalyst amount of 5 g the reaction rates approach a
Page 54
IDU -
140-()
„
— 120-"~-~.
>?'>
=o 10°-
cu
cuw
80-
.59.
c
2 60-CD
iffT3
1 40"
^->
C_
20-
0- 1 1 1 1 1 1 '
10 15
mass catalyst / [g]
20
Fig. 5-14: Initial differential selectivity using different amounts of catalyst. Reaction
conditions: 240 g BN, 100°C, 1 MPa and 1000 rpm.
1 6-
1 4-
1 2-
cu
Id 10-
o
=6 o8'
CD
çu"S O6'CD
Ë
04-
02-
00-
-— -d[BN]/dt / [mmol/s]-•— d[BA]/dt / [mmol/s]* d[DBA]/dt/[E-1 mmol/s]
-—d[DBI]/dt/[E-1 mmol/s]
10 15 20
mass catalyst / [g]
Fig. 5-15 : Initial reaction rates using different amounts of catalyst. Reaction conditions:
240 g BN, 100°C, 1 MPa and 1000 rpm.
Page 55
limiting value. This limit is given by the rate of hydrogen transfer to the
catalyst surface (see Chapter 5.3).
An explanation for the increasing selectivity towards butylamine using
higher catalyst loadings can be twofold: With an increasing amount of
catalyst per reaction volume the hydrogénation rate increases and thus also
the local temperature due to the exothermic reaction. This in turn accelerates
the reaction even more (runaway) and thus decreases the local hydrogen
concentration at the catalyst surface. These local temperature and hydrogen
concentration gradients are the larger the higher the catalyst concentration
is. The stirring is not efficient enough to cancel out these gradients. As a
consequence, a larger amount of catalyst per reaction volume decreases the
local hydrogen concentration and increases the local temperature. Both
effects result in an increase in selectivity towards butylamine.
Such a selectivity behavior is predicted by the approximate equation
(Eq. 5.10), mainly for the special cases I and III.
If the stirrer speed is risen from 1000 rpm to 1500 rpm the reaction rates
rose only slightly whereas the selectivity did not change. These results are
shown in Table 5-2 (the conditions were: 240 g BN, 3.75 g RaNi, 100°C and
Table 5-2: Selectivity and reaction rates at different stirrer speed. Reaction conditions:
240 g BN, 3.75 g RaNi, 100°C and 1 MPa.
stirrer speed [BA] [DBA] [max. DBI] -d[BN]/dt d[BA]/dt
/ [rpm] / [mass-%] / [mass-%] / [mass-%] / [mmol/(s*kg)] / [mmol/(s*kg)]
1000 88.6 11.0 9.95 118 83
1500 88.7 10.6 10.6 131 87
1 MPa). This is a second indication that no gas liquid transfer limitation for
hydrogen exists at the following conditions: 240 g BN, 3.75 g RaNi, 100°C,
1 MPa and 1000 rpm. An important note is, that this conditions are near the
gas-liquid transfer limitation for hydrogen.
Page 56
5.4.4 Recycling of the catalyst
Experiments to investigate the catalyst deactivation were carried out at the
following conditions: 240 g BN, 3.75 g RaNi, 100°C, 1 MPa and 1000 rpm.
The results are presented in Figure 5-16 Figure 5-17, Figure 5-18. The
—— butylamine
—•—dibutylamine* tributylamidine
—— max. dibutylimine
12 3 4
cycle / [-]
Fig. 5-16: Product distribution of butyronitrile hydrogénation recycling the same
catalyst four times. Reaction conditions: 240 g BN, 3.75 g RaNi, 100°C, 1
MPa and 1000 rpm.
selectivity towards the primary amine as well as the activity and the initial
differential selectivity are decreasing. In both cases, the decrease is large
from the first to the second cycle and small in the following cycles. Due to
this decrease in activity it follows that the local gradients of temperature and
hydrogen concentration are becoming less and less pronounced (see
Chapter 5.3). Following what has been said in Chapter 5.4.3 about the
effects of these local gradients one must expect a decrease in selectivity with
increasing the catalyst cycles. Such a selectivity behavior is predicted by the
approximate equation (Eq. 5.10) mainly for the special cases II and IV.
Page 57
60-
50-
40-
0)
"cd
g"G
S 30-
i_
nj
c 20-
10-
-d[BN]/dt / [mmol/(s*kg)]
d[BA]/dt / [mmol/(s*kg)]
d[DBA]/dt/[E-1 mmol/(s*kg)]
d[DBI]/dt/[E-1 mmol/(s*kg)]
cycle / [-]
Fig. 5-17: Initial rates of butyronitrile hydrogénation recycling the same catalysts four
times. Reaction conditions: 240 g BN, 3.75 g RaNi, 100°C, 1 MPa and
1000 rpm.
18-1
16-
14-1
^ .
>s
>12-
ocu
CD10-
CO.
CD8-
c
fl) .
l_
ë R-
T3 -
CD 4-
C-
2-
0-
cycle / [-]
Fig. 5-18 : Initial differential selectivity recycling the same catalysts four times. Reaction
conditions: 240 g BN, 3.75 g RaNi, 100°C, 1 MPa and 1000 rpm.
Page 58
5.4.5 Influence of additives
The influence of additives on the selectivity and activity was investigated
using standard conditions (240 g BN, 3.75 g RaNi, 100°C, 1 MPa, 1000
rpm). The results are presented in Figure 5-19 and Figure 5-20. The
100-
95-
II butylamine
liiilll dibutylamine
H tributylamidine
I 1 max. dibutylimine
additive
Fig. 5-19: Influence of additives on the product distribution. Reaction conditions: 240 g
BN, 3.75 g RaNi, 100 °C, 1 MPa and 1000 rpm.
following additives were tested: EDS, pyridine and ammonia. If EDS is
used, the reaction rate as well as the selectivity towards butylamine are
decreasing. EDS is known as a poison for hydrogénation catalysts. An
explanation for this results is, that EDS is poisoning the hydrogénation sites
of the catalyst and thereby slowing down the hydrogénation (-d[BN]/dt and
d[BA]/dt). Because EDS does not poison the acid sites, the side reactions
(BI DBI) are not slowed down and the measured selectivity towards
butylamine is lowered. Addition of pyridine does neither change the
selectivity nor the reaction rates significantly. If ammonia is added, the
selectivity towards butylamine is increased. The hydrogénation reactions are
Page 59
220-1
200-
180-
160-
\ZZ\ -d[BN]/dt/[mmol/(s*kg)]
^M d[BA]/dt/[mmol/(s*kg)]
^M d[DBA]/dt/[E-1 mmol/(s*kg)]
i I d[DBI]/dt/[E-1 mmol/(s*kg)]
,«PS
additive
Z1*«V' ai.6^w' 9^X A^^,d^e
5<â^
Fig. 5-20: Influence of additives on the initial rates of butyronitrile hydrogénation. Re¬
action conditions: 240 g BN, 3.75 g RaNi, 100 °C, 1 MPa and 1000 rpm.
slowed down, but not as dramatically as the side reactions (BI DBI).
The maximum intermediate concentration of dibutylimine is reduced too.
Explanations for this behavior are given in Chapter 4.2.3.
5.5 Influence of washing/modification procedures
5.5.1 Influence of washing procedures with different solvents
In order to describe the changes of catalyst properties brought about by
various washing procedures, the washed catalysts were tested using the
butyronitrile hydrogénation reaction. Thereby, standard conditions were
used: 240 g BN, 3.75 g RaNi, 100°C, 1 MPa, 1000 rpm. The results are
plotted in Figure 5-21 and Figure 5-22. The water treatment does neither
change the selectivity towards the primary amine nor the reaction rates.
Page 60
105-
100-
^
95-
butylamine
dibutylamine
tributylamidine
max. dibutylimine
fresh 3xH20 3xMeOH 3xEtOH 5% formaldehyde
Fig. 5-21: Influence of the washing procedure on the product distribution. Reaction
conditions: 240 g BN, 3.75 g RaNi, 100 °C, 1 MPa and 1000 rpm.
-d[BN]/dt / [mmol/(s*kg)]
d[BA]/dt/[mmol/(s*kg)]
d[DBA]/dt / [E-1 mmol/(s*kg)]
d[DBI]/dt/[E-1 mmol/(s*kg)]
fresh 3xH20 3xMeOH 3xEtOH 5% formaldehyde
Fig. 5-22: Influence of the washing procedure on the initial rates of butyronitrile
hydrogénation. Reaction conditions: 240 g BN, 3.75 g RaNi, 100 °C, 1 MPa
and 1000 rpm.
Page 61
Washing the fresh catalyst with methanol or modify it using formaldehyde
is dramatically decreasing (by a factor of 10) the reaction rates. The
selectivity towards butylamine is also decreasing. This is not in agreement
with a published patent of Degischer and Rössler. An explanation for this
behavior is that at the given conditions, the hydrogénation rates are so slow,
that the formation of secondary amine from butylamine already present in
the reaction mixture is lowering the yield of butylamine.
The influence of the washing procedure was also investigated under
conditions of gas-liquid transfer limitation for hydrogen (compare
Chapter 5.4.3). The following conditions were used: 240 g BN, 15 g RaNi,
100°C, 1 MPa and lOOOrpm. The results are plotted in Figure 5-23 and
110
105
100
£ 95
E 85
1 80
-g 15o
^10
5
0
fresh 3 x H20 3 x MeOH 5% formaldehyde
Fig. 5-23: Influence of the modification on the product distribution at conditions of a
hydrogen transfer limitation. Reaction conditions: 240 g BN, 15 g RaNi, 100
°C, 1 MPa and 1000 rpm.
Figure 5-24. Again the water washing does not change the selectivity and
only a small change in the reaction rates was measured. The reaction rates
for the methanol washed and the formaldehyde modified catalyst decrease
Page 62
butylamine
dibutylamine
tributylamidine
dibutylamide
fresh 3xH20 1xH20,3xMeOH 5% formaldehyde
Fig. 5-24: Influence of the modification on the initial rates of butyronitrile
hydrogénation at conditions of a hydrogen transfer limitation. Reaction
conditions: 240 g BN, 15 g RaNi, 100 °C, 1 MPa and 1000 rpm.
again, but the decrease is not as dramatically as if no hydrogen transfer
limitation exists, or in other words, the reaction rates using the unmodified
catalyst are slowed down because of the hydrogen transfer limitation. As
expected, the selectivity towards the primary amine is rising due to the local
hydrogen concentration (see Chapter 5.4.1) if high amounts of catalyst are
used.
5.5.2 Influence of the modification by formaldehyde
The influence of the strength of the formaldehyde treatment (compare
Chapter 6) was investigated using the conditions outside the transfer
limitation region for hydrogen. The selectivity towards butylamine was
higher if a low concentration of formaldehyde was used to modify the
catalyst (from 88.6 mass-% to 93.3 mass-%, compare Figure 5-25). The
hydrogénation rates decrease dramatically, so that if higher formaldehyde
concentrations are used to modify the catalyst, the hydrogénation reactions
Page 63
100-1
butylamine
dibutylamine
Bli tributylamidine
I 1 max. dibutylimine
unmodified 1% 2% 3.5%
[ChLO] / [mass-%]
Fig. 5-25 : Influence of the concentration of formaldehyde used to modify Raney nickel
on the product distribution. Reaction conditions: 240 g BN, 3.75 g RaNi,
100°C, 1 MPa and 1000 rpm.
are so slow, that the dimerisation can compete successfully the
hydrogénation (Table 5-3 and Figure 5-26). In Figure 5-26 the relative
initial reaction rates that were calculated by dividing the initial rates
obtained with the modified catalyst by the initial rates of the unmodified
catalyst are compared for various modification procedures. The reaction
leading to DBI (d[DBI]/dt) is somewhat stronger decelerated than the
reactions leading to butylamine.
Figure 5-27 reveals that the measured data points for the modified
catalysts do not fit the correlation between the yield butylamine versus the
initial differential selectivity for the unmodified catalysts shown in
Figure 5-5. Thus, the improved selectivity behavior of the modified catalyst
can not be solely explained by mass and heat transport limitation effects.
If modified nickel-on-carrier is used at the same conditions, again a rise
in selectivity is observed (Figure 5-28). The benefit of the modification
(from 84.1 mass-% to 92.1 mass-%) is even more impressive as in the case
of Raney nickel. Decreased hydrogénation rates were observed again, if the
Page 64
catalyst was modified (Table 5-4 and Figure 5-29). Plotting the relative ini¬
tial reaction rates again shows a stronger deceleration of the reaction leading
to dibutylimine.
Table 5-3: Influence of the modification strength on the initial reaction rates. Reaction
conditions: 240 g BN, 3.75 g RaNi, 100°C, 1 MPa and 1000 rpm.
modification
unmodified
1% CH20
2% CH20
3.5%CH20
5% CH20
-d[BN]/dt d[BA]/dt d[DBA]/dt d[DBI]/dt
/ [mmol/s*kg] / [mmol/s*kg] / [mmol/s*kg] / [mmol/s*kg]
118.656
82.262
42.310
18.743
12.233
82.612
59.926
31.967
13.771
8.806
3.199
2.322
1.034
0.455
0.314
14.258
8.194
3.633
1.835
1.276
-d[BN]/dt / [118.656 mmol/(s*kg)]
d[BA]/dt / [82.612 mmol/(s*kg)]
d[DBA]/dt/ [3.199 mmol/(s*kg)]
d[DBI]/dt / [14.258 mmol/(s*kg)]
unmodified 1% 2% 3.5%
[CH20] / [mass-%]
5%
Fig. 5-26: Influence of the formaldehyde modification of Raney nickel on the relative
initial reaction rates. Reaction conditions: 240 g BN, 3.75 g RaNi, 100°C,
1 MPa and 1000 rpm.
Page 65
94-
92-
toau¬
to03
"88-
CD ^ various reaction
CDCO
!LUB|Aj parameters (see Fig 6-3)
.q 84- • 1%CH202% CH203.5% CH205% CH20
o
CD
's* 82-
80-
1 '
10
1 '
20
1
30
i ' i
40 50
initial differential selectivity / [-]
Fig. 5-27: Initial differential selectivity versus product distribution (yield butylamine).
100 -,
95-
<n
E
o
'l_
W
T3
T3
O
E3
butylamine
dibutylamine
tributylamidine
max. dibutylimine
unmodified 2.5%
[CKO] / [mass-%]
Fig. 5-28: Influence of the formaldehyde modification of nickel-on-carrier on the
product distribution. Reaction conditions: 240 g BN, 10 g nickel-on-carrier,
100°C, 1 MPa and 1000 rpm.
Page 66
Table 5-4: Influence of the formaldehyde modification of nickel-on-carrier on the initial
reaction rates. Reaction conditions: 240 g BN, 10 g nickel-on-carrier, 100°C,
1 MPa and 1000 rpm.
modification -d[BN]/dt d[BA]/dt d[DBA]/dt d[DBI]/dt
/ [mmol/s*kg] / [mmol/s* kg] / [mmol/s*kg] / [mmol/s*kg]
unmodified 30.605 15.979 1.430 6.219
2.5% CH20 14.867 10.673 0.448 1.542
5% CH20 8.810 5.734 0.256 1.256
unmodified
-d[BN]/dt/ [30.605 mmol/(s*kg)]
d[BA]/dt / [15.979 mmol/(s*kg)]
d[DBA]/dt / [1.430 mmol/(s*kg)]
d[DBI]/dt/ [6.219 mmol/(s*kg)]
2.5%
[CH O] / [mass-%]
Fig. 5-29: Influence of the formaldehyde modification of nickel-on-carrier on the
relative initial reaction rates. Reaction conditions: 240 g BN, 10 g nickel-on-
carrier, 100°C, 1 MPa and 1000 rpm.
Page 67
5.6 Reversibility of the hydrogénation steps
5.6.1 Disproportionation of butylamine
To investigate the reversibility of the hydrogénation steps, experiments
using butylamine and dibutylimine as starting materials were carried out. In
a first experiment, the conditions were identical with those used as standard
conditions to hydrogenate butyronitrile (254 g BA, 24 g RaNi, 100°C,
1 MPa, and 1000 rpm). The reaction profile is shown in Figure 5-30. The
100
90
80
^ 70
COCO 60cc
F50
c
o 40
CO
oQ.
30
EoÜ
20
10
butylamine
dibutylamine
tributylamine
~i
10
time / [day]
Fig. 5-30: Reaction profile using butylamine as starting material. Reaction conditions:
254 g BA, 24 g RaNi, 100 °C, 1 MPa and 1000 rpm.
first derivation of the concentration profile results in the disproportionation
rates of the primary amine at a given concentration. Using this experiment,
the reversibility of the hydrogénation reaction leading from the primary
amine to the secondary amine is proved. This experiment also shows that
the amount of the primary amine obtained by the catalytic hydrogénation of
Page 68
butyronitrile with Raney nickel is mainly determined by the kinetics and not
by the thermodynamics (equilibrium) of the process. This fact may also
explain the obtained low selectivity of the third cycle when the recycling of
the catalyst was investigated (Figure 5-16) or the low selectivity of the
catalyst modified by 5% formaldehyde. If the hydrogénation is to slow, i.e.
if the reaction time is to high, the obtained selectivity towards butylamine is
not the maximum possible one (kinetic) but strives towards the much lower
thermodynamically determined product distribution. This is shown
in Figure 5-31.
100
80-
co
CO
| 60
w 40oQ.
EoÜ
20
—— butylamine
• butyronitrile* dibutylimine
——dibutylamine
WW*---
>- » » I—•" 1 //—100 200 300 1000
time / [min]
1100 1200
Fig. 5-31: Hydrogénation of butyronitrile (not stopped after full conversion). Reaction
conditions: 240 g BN, 3.75 g RaNi, 100 °C, 1 MPa and 1000 rpm.
A second experiment was performed, using dibutylamine as starting
material. 240g DBA, 20 g RaNi and 30 g NH3 were placed in the reactor.
The stirrer speed was set at 1000 rpm and the temperature was 100°C. No
reaction was observed, after 5000 min. The explanation for this behavior is,
that the dibutylamine could not make sufficient contact with the catalytic
surface of the wet catalyst.
Page 69
5.6.2 Influence of reaction parameters on the disproportionation of
butylamine
The influence ofthe temperature and the hydrogen pressure was investigated
using the 35 ml reactors (see Chapter 9.1.4) and the procedure described in
Chapter 9.2.2. The reactors were charged with high catalyst loadings, and
the hydrogen was pressed into the reactor before heating. The pressure
during the reaction was not recorded. It is evident, that the temperature has a
great influence on the disproportionation rate (Figure 5-32).
— 100 °C
• 120°C
* 140°C
——160°C
U-| 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0
time / [day]
Fig. 5-32 : Influence ofthe temperature on the disproportionation rate ofbutylamine. Re¬
action conditions: 20 g BA, 2 g RaNi, 280 min."1 without hydrogen.
The hydrogen pressure also influences the disproportionation rate.
Higher hydrogen pressures decrease the reaction to the secondary amine
(Figure 5-33). A dehydrogenation of butylamine as the first step in the
disproportionation mechanism could be an explanation for this behavior [41].
As mentioned above, a faster disproportionation of butylamine can
decrease the high yield obtained under kinetic controlled conditions. This
Page 70
110-1
wCO
CD
E
cu
g
ETO
.Q
without H„
2 MPa H24 MPa H„
5 6
time /[day]
11
Fig. 5-33: Influence of the hydrogen pressure on the disproportionation of butylamine.
Reaction conditions: 20 g BA, 2 g RaNi, 120°C and 280 min."1.
CO
CO
CD
E
110-,
100-
90-
80-
70-
60-
cu
ETO
X2
50-
40-
30-
20-
10-
-0 1
unmodified
1%CH20 modified
2% CH20 modified
3.5% CH20 modified)
5% CH20 modified
00
—r~
01
—r~
02 03
time / [day]
—i—
04 05
—I—
06
—I
07
Fig. 5-34: Influence of the catalyst modification on the disproportionation of butyl¬
amine. Reaction conditions: 20 g BA, 2 g RaNi, 120°C and 280 min."1.
Page 71
argumentation was also used to explain the lower selectivity, if a catalyst,
modified by 5% formaldehyde was used. The influence of the catalyst
modification by formaldehyde on the disproportionation rate is shown
Figure 5-34.
5.6.3 Dibutylimine as starting material
Dibutylimine was synthesised from butylamine and butyraldehyde using an
excess of butylamine. 174 g of the obtained solution, containing 69.5%
dibutylimine (0.95 mol), 12.4% butylamine and 18.1% side products from
the synthesis was used for the disproportionation study. The reaction
conditions were: 3.79 g RaNi, 20 g NH3 (1.18 mol), 100°C, 2 MPa,
1000 rpm. During the heating period dibutylimine already disappears and
butyronitrile and butylamine are produced This experiment shows that the
hydrogénation of butyronitrile is reversible too.. A second result is, that the
COCO
CD
E
o
"55oQ.
Eoo
90-1
80-
70-
60-
50-
40-
30-
20-
10-
0-
-— butylamine-•— butyronitrile*— dibutylimine-— dibutylamine
side products
^ y w w ^ w'—iip • |p ~— up ~— up —-
-50
T
50 100 150 200 250 300 350 400
time / [min]
Fig. 5-35: Reaction profile using dibutylimine as starting material. Reaction conditions:
174 g DBI, 3.75 g RaNi, 20 g ammonia, 100°C, 2 MPa and 1000 rpm
Page 72
amount of dibutylimine is not decisively influencing the selectivity of this
reaction, or in other words: the reaction rates transforming dibutylimine,
butylamine and butylimine to the state of equilibrium are faster than the
reaction rate of the hydrogénation of butylimine leading to butylamine
(Figure 5-35).
5.7 Discussion
5.7.1 The bifunctional catalytic hydrogénation and its reversibility
The mechanism presented in Chapter 3.1 (Figure 3-1) postulates, that the
selectivity of the hydrogénation of nitriles depends on two catalytic
functions, an acidic and a hydrogénation function. The ratio and the distance
between these two functions is assumed to determine the ratio of the
reaction constants kH4/kC5 (in Figure 3-1) and therefore the yield of primary
amine. Furthermore, the maximum produced amount of dibutylimine was
observed to be higher than the produced amount of dibutylamine at full
conversion, so that a reaction path from dibutylimine to butylamine is
highly probable (see Figure 5-9, Figure 5-13 and Figure 5-31). From the
experiments using dibutylimine and ammonia as starting material it is
obvious, that high selectivities towards the primary amine can be obtained,
even if the condensation products are already present in the reaction
mixture. This leads to the conclusion, that dibutylimine is in a fast
equilibrium with ammonia and a species that can be hydrogenated to the
primary amine. In such disproportionation experiments with the butylamine
or the condensation by-products as starting materials butyronitrile as
intermediate is detected. This reveals the reversibility of the hydrogénation
as well as the condensation steps. Depending on the reaction conditions
applied, such reversible reaction systems can exhibit, as two extremes, a
kinetically or a thermodynamically determined product distribution. The
catalytic hydrogénation of the aliphatic butyronitrile is obviously such a
reaction system with butylamine as the kinetically controlled product. Thus,
Page 73
in order to optimize the selectivity towards butylamine, the reaction must be
well monitored over time and immediately stopped after conversion has
been completed. If the reaction proceeds beyond this time the butylamine
disproportionates and the reaction strives towards a product distribution
determined by the equilibrium. This, however, is unfavourable for a
selective production of butylamine.
5.7.2 Influence of various reaction parameters on the selectivity
First attempts to get a qualitative understanding of the influence of various
reaction parameters on the selectivity towards the desired primary amine
failed due to the complexity of the hydrogénation system. Therefore, a
simple semi-quantitative macro-kinetic model has been derived, with the
help of which the complex selectivity behavior of the explored reaction
system could more easily be characterised. For the sake of simplicity the
rate constants of the adsorption and desorption steps of the various reaction
components to and from the catalyst surface, respectively, has in each case
been incorporated into the overall rate constant of the corresponding
reaction step. The explanation of the effects of various reaction parameters
such as the gas-liquid transfer limitation of hydrogen, the overall pressure,
the temperature, the amount and recycling of catalyst and the presence of
various additives on the yield of the desired butylamine led to the
conclusion that all measured parameter effects can be explained if the
following question can be answered: "What influence do the changes of the
various reaction parameters exert on the local gradients of temperature and
hydrogen concentration near the catalyst surface?" It has been shown
(Chapter 5.4) that the selectivity towards butylamine increases if the local
temperature increases (Chapter 5.4.2) or if the local hydrogen concentration
decreases (Chapter 5.4.1). This can be rationalised with the help of the
"quasi" Arrhenius plot (Figure 5-12) and the equation for the differential
selectivity (Eq. 5.10) taking also into account aspects of mass and heat
transfer effects (Chapter 5.3). The more these local gradients change by
changing the reaction parameters the lower the influence on the selectivity
Page 74
becomes. Therefore, the most favourable selectivity towards butylamine can
be expected, if highly active catalysts are used and as a consequence, a
temperature and mass diffusion limitation becomes dominant causing high
local gradients.
Page 75
Page 76
Chapter Vf
Modification of nickel catalysts by
formaldehyde
6.1 General remarks
The modification mechanism of nickel or cobalt catalysts by formaldehyde
is still unknown. The interaction of formaldehyde and/or products of its
decomposition or polymerisation with the catalyst, produces a catalyst with
different properties, thereby producing higher selectivities towards the
primary amines, if nitriles are hydrogenated [1, 63, 64, 81]. One explanation
for the interaction of formaldehyde and the nickel catalyst is a
disproportionation of formaldehyde to methanol and carbonmonoxide
(Eq. 6.1).
2 CH20 CH3OH + CO Eq. 6.1
The formed methanol desorbs from the catalyst while carbonmonoxide is
still adsorbed on the catalysts surface as Nix(CO)y. This mechanism would
be in agreement with reported results, where the interaction of
formaldehyde with Ni(llO) surfaces was investigated at low temperatures
(95 K) by Richter and Ho [82]. Formaldehyde reacts on the catalysts
surface, producing a mixed CH30 and CO adlayer on the surface (Eq. 6.2).
CH20 + CH20 CH30(a) + CO(a) + H(a) Eq. 6.2
The adsorbed species that arised due to the adsorption and thermal
processing of CH20 on Ni(llO) were summarised in Eq. 6.3-Eq. 6.7. Mixed
Page 77
paraformaldehyde and solid H2CO multilayers were formed at high
formaldehyde exposures.
CH20(g) CO(a) + 2 H(a) Eq. 6.3
2 CH20(g) CH30(a) + CO(a) + H(a) Eq. 6.4
n CH20(g) (CH20)n(a) Eq. 6.5
(CH20)n(a) HCOO(a) + CH30(a) + CO(a) + H(a) Eq. 6.6
+ CHx(a) + CH20(g)
CH30(a) *- CO(a) + 3 H(a) Eq. 6.7
If the catalyst was heated after the formaldehyde treatment, methanol
desorbed from the catalyst at temperatures of about -10°C, carbonmonoxide
desorbed at temperatures above 174°C.
Newton and Dodge [83] investigated the equilibrium constants between
carbon monoxide, hydrogen, formaldehyde and methanol (Eq. 6.8 and
Eq. 6.9).
CO + H2 ^=^ CH20 Eq. 6.8
CH20 + H2 =*=^ CH3OH Eq. 6.9
The scope of their work was to investigate, if it is possible to produce
formaldehyde by hydrogénation of carbon monoxide. One of the results
was, that nickel catalysts promote the decomposition of formaldehyde into
carbon monoxide and hydrogen. At 200°C and 0.1 MPa the equilibrium
constants were Kj = 2.30 * 10"5 (Eq. 6.8) and K2 = 1800 (Eq. 6.9). The
conclusion was, that the production of formaldehyde from carbon monoxide
by hydrogénation is not feasible at any reasonable temperature or pressure.
Several investigations concerning the chemisorption of CO on different
Ni surfaces were recently made [84-87]. Especially the adsorption mode
Page 78
was investigated, thereby two species were observed: Bridged CO and
on-top or terminal CO. Both species coexist and are in an equilibrium
depending on pressure (surface coverage) and temperature. The desorption
temperature is in the range of 170°C.
The washing operations as well as the formaldehyde treatment were
made in a three-neck sulfonation flask (see Chapter 9.1.5) as described in
Chapter 9.2.7-9.2.10.
Modification experiments were usually made, using the procedures of
Degischer and Rössler [1, 63]. A 5% formaldehyde modification of a Raney
catalyst means, that X g Raney nickel are modified by 2X g of an aqueous
solution containing 5% formaldehyde (normal procedure: 50 g Raney nickel
in 100 g aqueous solution, that contains 5 g formaldehyde).
Methanol (by headspace-GC), formaldehyde (by HPLC) and the nickel
concentration (by XRF) were determined in the aqueous modification
solution after a modification time of 30 min.
6.2 Influence of the treatment with different solvents on the
properties of the catalyst
6.2.1 Reduction potential
The standard reduction potential of fresh, water washed, alcohol and
formaldehyde modified catalysts was measured as described in
Chapter 9.5.1. The potential of the fresh and the water washed catalyst is in
a range of -0.6 V indicating that the water washing does not change the
properties of the catalyst. This observation is in agreement with the results
of the nitrile hydrogénation (see Figure 5-19 and Figure 5-20). The
potential of the methanol, ethanol and formaldehyde modified catalyst is in
a range of -0.3 V, indicating that these substances do change the catalysts
properties. In addition, if the fresh catalyst was washed with water or
methanol, no nickel was found in the washing solutions.
Page 79
650-
600-
550-
^ 500-
£ 450-
öj 400-
S 350-
Q. 300-
1 250-
T3c 200-
3« 150-
I
100-
50-
0- -1 '—'—' 1 '—'—' 1 '—'—' 1-
fresh 3xH20 3xMeOH 3xEtOH 5% formaldehyde
Fig. 6-1: Standard reduction potential of fresh and modified catalysts determined with
a combined gold electrode.
6.2.2 Adsorption of an indicator
In order to get some information about the influence of the modification
treatments on the acid sites activity the adsorption capacity for
4-aminoazobenzene was determined for fresh, water washed, methanol and
formaldehyde modified catalysts as described in Chapter 9.5.2. Four
measurements were carried out. Figure 6-2 reveals that the acid activity of
the catalysts does depend on the modification treatment applied. However,
no simple correlation between these measurements and the selectivity data
was found.
Page 80
40-
35-
, ,
n>
o 30-
F^r
"
O2 b-
'—'
—..
>s
75 20-
roQ.
mo 1 5-
c
o-!—«
Q. 1 0-
O<n
"
n
ro 05-
00-
fresh 3 x H20 3 x MeOH 5% formaldehyde
Fig. 6-2 : Adsorption of 4-aminoazobenzene on Raney nickel.
6.3 Modification of Raney nickel by various formaldehydeconcentrations
6.3.1 Analysis of the modifying solution
Different formaldehyde concentrations were used (1%, 2%, 3.5%, 5% and
7% in water), while the mass of catalyst and the mass of the aqueous
solution were not varied (50 g wet Raney nickel and 100 g solution). The
methanol, formaldehyde (which was not converted during the modification)
and residual formaldehyde amounts are plotted in Figure 6-3. At low
formaldehyde concentrations, formaldehyde disappears completely, at
higher concentrations not all formaldehyde is consumed. A remarkable
result is, that the amount of methanol corresponds to the half amount of the
consumed formaldehyde at low modification concentration (Table 6-1). In
this table, the mass balance of the modification procedure is given, as well
as the division of methanol found by the residual mass in [mmol CH20] that
was not found. At low concentrations, the stoichiometry of methanol and
Page 81
[CH.O] / [mass-%]
Fig. 6-3: Amount of methanol, formaldehyde and residual mass in the modifying
solution at various modification conditions. Reaction conditions: 50 g Raney
nickel modified in total 100 g aqueous solution.
Table 6-1: Mass balance of the catalyst modification by various formaldehyde
concentrations. Initial formaldehyde, not converted formaldehyde and
methanol determined in the modification solution and the residual mass
(volume: 100 ml).
initial CH20 MeOH residual MeOH/ hydrogen needed
CH20 measured measured residual to produce MeOH
/ [mmol] / [mmol] / [mmol] / [mmol CH20] /[-] / [ml/g cat]
33.9 0.0 17.0 17.0 1.00 7.6
68.0 0.0 29.4 38.6 0.76 13.2
118.1 5.9 51.6 60.7 0.85 23.1
169.2 11.7 87.3 70.2 1.24 39.5
237.0 66.4 87.6 83.0 1.06 39.2
Page 82
residual mass is 1, indicating that formaldehyde reacts according Eq. 6.1 to
methanol and carbon monoxide. The last row of the table, describes the
hydrogen, that would be needed to produce the measured methanol from
formaldehyde by direct hydrogénation. This value was 39 ml hydrogen per
g catalyst at high modification concentrations. Literature values for the
hydrogen content stored on Raney nickel are about 20 ml/g, so that not
enough hydrogen is stored on the catalyst to produce the methanol measured
in the modification solution [45].
The nickel concentration in the liquid phase after modification is shown
in Figure 6-4. The higher the formaldehyde concentration is, the higher is
the nickel amount removed from the catalyst. There exists almost a linear
correlation between the modification concentration and the leached nickel.
5000-
4000-
EQ.
B: 3000•
o
^ 2000-
CD.co
CD
_ÇD
1000 4
[CH20] / [mass-%]
Fig. 6-4: Concentration of leached nickel in the modifying solution determined by
XRF. Reaction conditions: 50 g Raney nickel in total 50 g aqueous solution.
Page 83
6.3.2 Properties of the modified catalysts
The standard reduction potential of the modified catalysts are presented in
Figure 6-5. The potential is higher, if the catalyst is modified by
600 -,
500-
•>-
E400-
-~~-
CD .
.|_«
c
CD 300-oQ.
T3i_
coT3
200-
C
m
w
100-
0 —'—I—! H-1 H-1 H 1—I—I H H 1 *—I—*
01 2345678
[CH20] / [mass-%]
Fig. 6-5: Standard reduction potential of Raney nickel modified by various
formaldehyde concentrations determined with a combined gold electrode.
formaldehyde. A value of -0.35 V is obtained for catalyst, treated by high
formaldehyde concentrations.
The stability against acidic attack is shown in Figure 6-6. The fresh
catalyst is completely dissolved in hydrochloric solution, while the 7.5%
formaldehyde modified catalyst at these conditions is stable and almost
releases no hydrogen.
Page 84
350-,
1
300-
(/)
>s -
m
m 250-o
n) -
E 200-
—.
T3"
CD
ce150-
fl)
CDi_
c 100-CDO)ol_
T3 50-
.c
°H—>—i—'—i—'—i—'—i—'—i—'—i—'—i—'—i01 2345678
[CH20] / [mass-%]
Fig. 6-6: Hydrogen evolved by treating Raney nickel catalysts with hydrochloric acid.
Reaction conditions: 1 g Raney nickel was dissolved in 70 g of a 10%
hydrochloric acid.
6.4 Modification of various amounts of Raney nickel at
constant modification strength
6.4.1 Analysis of the modifying solution
Different catalyst amounts (12.5 g, 25 g, 37.5 g, 50 g, 75 g and 100 g) were
used while the formaldehyde concentration and the mass of the aqueous
solution were constant (113 g of a 4.5% formaldehyde solution). The
methanol and residual formaldehyde amounts are plotted in Figure 6-7. The
amount of disappeared formaldehyde, that is not found as methanol in the
liquid phase is plotted as residual mass. At low catalyst amounts, not all the
formaldehyde reacts. At higher catalyst amounts, all formaldehyde
disappears. The same calculation as in Chapter 6.3 was made again and the
results are listed in Table 6-2. At high catalyst masses (low formaldehyde
concentration / catalyst ratio), the stoichiometry between methanol found in
Page 85
40 60 80 100
amount catalyst in a 4.5% formaldehyde solution / [g]
Fig. 6-7: Amount of methanol, formaldehyde and residual mass found in the modifying
solution. Reaction conditions: 113 g of an aqueous solution containing 4.5%
formaldehyde.
Table 6-2: Mass balance of the catalyst modification by various catalyst amounts. Mass
of catalyst used, not converted formaldehyde and methanol determined in the
modification solution and the residual mass (initial formaldehyde amount
169.2 mmol, volume: 113 ml).
catalyst CH20 MeOH residual MeOH/ hydrogen needed
amount measured measured residual to produce MeOH
/[g] / [mmol] / [mmol] / [mmol CH20] /[-] / [ml/g cat]
12.5 119.2 14.2 35.9 0.39 25.1
25 92.9 30.8 45.4 0.68 27.2
37.5 47.7 55.1 66.5 0.83 33.0
50 49.6 57.2 62.6 0.91 25.5
75 0 86.5 82.7 1.05 25.84
100 0 86.9 82.5 1.05 19.5
Page 86
the solution and residual mass is almost 1, at low amounts of catalyst, the
stoichiometry is far from 1.
The nickel concentration in the liquid phase after modification is shown
in Figure 6-8. The more catalyst being present in the modifying system, the
more nickel is leached from the catalyst.
3000-
2500-
E 2000-Q.Q.
CD
ü
oCD
Ücc
CD
1500-
1000-
500-
20 40 60 80 100
amount catalyst in a 4.5% formaldehyde solution / [g]
Fig. 6-8: Concentration of leached nickel in the modifying solution determined by
XRF. Reaction conditions: 113 g of an aqueous solution containing 4.5%
formaldehyde modifying various amounts of catalyst.
6.4.2 Properties of the modified catalysts
The standard reduction potential of the modified catalysts is plotted in
Figure 6-9. The amount of hydrogen which is released if the catalyst is
attacked by hydrochloric acid is shown in Figure 6-10. If the formaldehyde
modification is carried out with small amounts of catalyst the released
hydrogen remains the same.
Page 87
400-
350-
300-
rö 250 4
CD
"5 200-Q.
T3
ro lOO-
I 100.
50-
+- -+-
20 40 60 80 100
amount catalyst in a 4.5% formaldehyde solution / [g]
Fig. 6-9: Standard reduction potential of various catalyst amounts modified by a 4.5%
formaldehyde solution.
180-
160-
CO>< 140cch—»
cco 120O)
E100
^
c
CD
O)80
O&_
T>
>< HO.C
oCDCO 40
CCCD
P 20-
2080
40 6080
100
amount catalyst in a 4.5% formaldehyde solution / [g]
Fig. 6-10: Hydrogen evolved from differently modified catalysts after treatment with
hydrochloric acid. Reaction conditions: 1 g Raney nickel dissolved in 70 g of
a 10%) hydrochloric acid.
Page 88
6.5 Modification of nickel-on-carrier
The modification procedure was not only investigated using Raney
catalysts, but also using a nickel-on-carrier as hydrogénation catalyst. If the
catalyst was modified by a 5% formaldehyde solution (169 mmol
formaldehyde), not all formaldehyde was consumed (Table 6-3). If
unmodified catalyst is washed with water, no nickel is leached from the
catalyst (< 10 ppm). However, if the catalyst was modified by
formaldehyde, a concentration of 463 ppm nickel was found in the
modification solution.
Table 6-3 : Mass balance of the modification of nickel-on-carrier. Mass of catalyst used,
formaldehyde and methanol determined in the modification solution and the
residual mass.
catalyst
amount
/[g]
CH20
measured
/ [mmol]
MeOH
measured
/ [mmol]
residual
/ [mmol CH20]
MeOH/
residual
/[-]
hydrogen needed
to produce MeOH
/ [ml / g cat]
50 113.3 19.8 36.0 0.55 8.8
6.6 Discussion
The first conclusion which can be drawn from modification experiments
with formaldehyde is that changes of the catalyst properties due to water
washing operations before and after the modification by formaldehyde can
be excluded.
Up to 50% of the formaldehyde used to modify the catalyst can be
found as methanol in the solution. If modification conditions were used
whereby more than 50 g catalyst amount and a formaldehyde concentration
of 1 mass-% were applied almost exactly 50% were found as methanol. This
is an indication that the reaction of formaldehyde to methanol together with
Page 89
an adsorbed species occurs. The catalyst does not store enough hydrogen to
hydrogenate formaldehyde to the produced amount methanol. Therefore,
this additional hydrogen has to be taken from somewhere else presumedly
from formaldehyde itself or from the solvent. These facts are strong
indications that formaldehyde reacts according to Eq. 6.1 at low
concentrations of formaldehyde and high amounts of catalyst. The
production of chemisorbed carbon monoxide would also explain the effect
of other modifiers, such as acetaldehyde, benzaldehyde, carbon dioxide,
acetone and carbon monoxide itself that were described in the patent [1].
From all these modifiers, carbon monoxide could be produced on the
surface of the catalyst.
The effect of chemisorbed carbon monoxide on the hydrogénation of
nitriles could be explained by stereochemical restrictions. Carbon monoxide
blocks metal sites on the catalyst, so that larger substrates (dialkylimines)
can not adsorb as easily as smaller substrates (alkylimines). Tributylamidine
was observed as by-product if modified catalysts were used to hydrogenate
butyronitrile. One explanation for the formation of this product, is that this
large molecule can not be hydrogenated to tributylamine at the chosen
conditions because of stereochemical reasons.
Large amounts of nickel can be found in the modification solution, an
important fact, if modifications have to be made on a larger scale. The
minimization of this nickel leaching during the modification procedure and
also during the hydrogénation reaction is evidently important for the
preparation of large amounts of catalyst.
The catalysts treated by high formaldehyde concentrations are almost
stable in hydrochloric acid, while a fresh catalyst is completely dissolved in
a short time.
Page 90
Chapter /
Effect of formaldehyde modified nickel
catalysts on other chemical systems
Formaldehyde modified nickel catalysts have revealed an unexpected and
positive selectivity effect on the hydrogénation of nitriles (Chapter 5.5). The
exact reason for this observation is not yet known and its investigation shall
be the aim of future research studies. In order to examine whether these new
catalysts also bring about such selectivity effects in other well known
hydrogénation processes, in this Chapter some typical hydrogénations were
screened.
7.1 Hydrogénation of crotonaldehyde
7.1.1 General remarks
The hydrogénation of oc,ß-unsaturated aldehydes is still a challenging field
of investigation. The desired oc,ß-unsaturated alcohol is not the
thermodynamic product, and therefore, the saturated aldehyde is preferably
formed. A change of the selectivity has to be achieved by a change of the
reaction rate constants of the competitive reactions and of the competitive
adsorption constants of the components. The catalytic system that produces
the highest selectivities are Pt/Ti02, Pt/Si02, doped by different transition
metals or modifiers [88-91]. Selectivities up to 50% were achieved if
crotonaldehyde was hydrogenated in ethanol. NiPt/Si02 catalysts were also
tested, but produced only small amounts of the desired alcohol [92].
Page 91
Selectivities up to 85% towards the unsaturated alcohol can be achieved
using bimetallic Ag/Si02 catalysts [93, 94] or Ru/Si02 catalysts [95, 96].
Investigations on the influence of the catalyst modification on the
hydrogénation of oc,ß-unsaturated aldehydes were performed in a system of
crotonaldehyde and ethanol as solvent. A scheme of the reaction system is
given in Figure 7-1. E/Z crotonaldehyde reacts to butanol over butanal or
over E/Z crotylalkohol as intermediate.
E/Z crotylalkohol
Fig. 7-1 : Hydrogénation of E/Z crotonaldehyde to butanol over the intermediates
butanal or E/Z crotylalkohol.
Experiments were made in a 500 ml steel hydrogenator (see
Chapter 9.1.1) according the procedure described in Chapter 9.2.3, and
samples were taken according the procedure in Chapter 9.2.6.
1,1-diethoxybutane was observed as by-product (Figure 7-2) or even as
main product, in cases that Raney nickel was modified by formaldehyde.
This product is formed in an acid catalysed side reaction via the semiacetal
by addition of the solvent and subsequent dehydration [90].
Page 92
.0. 1,1 -diethoxybutane
CL /
Fig. 7-2: 1,1-diethoxybutane was formed as by-product, if crotonaldehyde was
hydrogenated.
7.1.2 Test for a possible gas-liquid transfer limitation for hydrogen
To investigate whether a gas-liquid transfer limitation for hydrogen exists at
the chosen reaction conditions, the amount of catalyst was doubled (40 g
E/Z crotonaldehyde, 200 g ethanol, 30°C, 1 MPa overall pressure and
1000 rpm). As can be seen in Table 7-1 and Figure 7-3 the reaction rate
increases as the catalyst amount was raised, so that a transfer limitation can
be excluded.,
Table 7-1 : Influence of the catalyst amount on the initial reaction rates, if E/Z croton¬
aldehyde is hydrogenated. Reaction conditions: 40 g E/Z crotonaldehyde,
200 g ethanol, 30°C, 1 MPa and 1000 rpm.
catalyst initia disappearance rate of initial production rate of
amount crotonaldehyde butanol
/[g] / [mmol/s] / [mmol/s]
2 0.193 0.0024
4 0.395 0.0049
Page 93
100-,
COcoce
COoQ.
OÜ
75
time / [min]
150
Fig. 7-3: Influence of the catalyst amount on the hydrogénation rates. Reaction
conditions: 40 g crotonaldehyde, 200 g ethanol, 30°C, 1 MPa and 1000 rpm.
7.1.3 Influence of the formaldehyde modification ofRaney nickel on the
hydrogénation of crotonaldehyde
Modification of Raney nickel with formaldehyde leads to a lower
disappearance rate of crotonaldehyde and a lower production rate for
butanol (Figure 7-4 and Figure 7-5). Butanal is the only observed
intermediate, no crotylalkohol was observed during the reaction. If the
catalyst is modified using a 5% formaldehyde solution, condensations take
place and 1,1-diethoxybutane is observed as main product.
Page 94
COCOce
oü
100
80-
60-
unmodified
1 % CH202% CH203.5% CH20
S 40-cooQ.
20-
100 125 150 175 200 225 250 275 300
time / [min]
Fig. 7-4: Influence of the modification strength of the catalyst on the crotonaldehyde
disappearance rate. Reaction conditions: 40 g crotonaldehyde, 2 g RaNi,
30°C, 1 MPa and 1000 rpm.
20-1
18-
16-
-9-
o^ 14-
CO
CO
CO 12-
E -
10-
c
o
-I-»8-
CO .
o
Q. 6-
E -
o 4-o
—— unmodified, butanol
—•—-1%CH20 modified, butanol
—A—- 2% CH20 modified, butanol
——- 3.5% CH20 modified, 1,1-diethoxybutane- 3.5% CH20 modified, butanol
—+—- 5% CH20 modified, 1,1-diethoxybutaneX—- 5% CH20 modified, butanol
1 r^T
150 200
T—' 1—' 1 '—I—' 1—' 1 '—I—' 1
250 300 350 400 450 500 550 600
time / [min]
Fig. 7-5: Influence of the modification strength of the catalyst on the product
distribution and formation rate. Reaction conditions: 40 g crotonaldehyde, 2 g
RaNi, 30°C, 1 MPa and 1000 rpm.
Page 95
7.1.4 Influence of the formaldehyde modification of nickel-on-carrier
on the hydrogénation of crotonaldehyde
The influence of the modification by formaldehyde on the selectivity and
activity was also investigated using a nickel-on-carrier catalyst. The
behavior of the system was as in the case of Raney nickel (Figure 7-6). The
0 50 100 150 200 250 300 350 400
time / [min]
Fig. 7-6: Influence of the modification of a nickel-on-carrier catalyst on the reaction
rates. Reaction conditions: 40 g crotonaldehyde, 4 g nickel-on-carrier, 30°C,
1 MPa and 1000 rpm.
modification lowers the hydrogénation rate of crotonaldehyde if
formaldehyde concentrations above ca. 2.5% for the modification were
used. No 1,1-diethoxybutane was observed as by-product using this catalyst.
Page 96
7.2 Hydrogénation of l-bromo-4-nitrobenzene
7.2.1 General aspects
Several possibilities exist to reduce nitroaromatic compounds. The
Béchamps-reduction, the reduction using other metals than iron, the
reduction by sulphides, electrochemical reductions or the reduction using
hydrazine [97].
The hydrogénation of nitroarenes, especially of halide substituted
arènes using Raney nickel as catalyst is of industrial interest [98]. One of the
problems thereby is the dehalogenation of the substrate or product.
Therefore, there was some hope that the newly discovered formaldehyde
nitrobenzene
Fig. 7-7: Reaction scheme for the hydrogénation of l-bromo-4-nitrobenzene to
l-bromo-4-aminobenzene and the undesired side-reactions (hydrogenolysis
of bromine) to aniline.
Page 97
modified catalyst might prevent the undesired dehalogenation. The
hydrogénation of a halide substituted nitroarene was investigated using the
substrate l-bromo-4-nitrobenzene. A reaction scheme is given in
Figure 7-7. The experiments were made in a 200 ml glass hydrogenator
(Chapter 9.1.2) according the procedure described in Chapter 9.2.4. After a
modification procedure the catalyst was washed three times with
tetrahydrofurane (Chapter 9.2.9). The reaction conditions were: 0.5 g Raney
nickel, 10 g l-bromo-4-nitrobenzene, 90 g THF, 50°C, 0.5 MPa and 1500
rpm.
7.2.2 Test for a possible gas-liquid transfer limitation for hydrogen
To check whether an influence of the gas-liquid transfer limitation for
hydrogen exists, the hydrogénation rate was doubled by using twice the
amount of the catalyst (Table 7-2). This is an indication that no limitation
Table 7-2: Selectivity and rate of the hydrogénation of l-bromo-4-nitrobenzene.
Reaction conditions: 10 g l-bromo-4-nitrobenzene, 90 g THF, 50°C, 0.5 MPa
and 1500 rpm.
Raneynickel initial hydrogen initial l-bromo-4- aniline
amount uptake hydrogénation rate aminobenzene
/[g] / [mmol/s] / [mmol/(s*kg)] / [mass-%] / [mass-%]
0.48 0.011 22.2 99.3 0.7
1.00 0.028 28.0 98.5 1.5
for hydrogen exists. In addition, if the catalyst was not washed with
tetrahydrofurane, no reproducible reaction rates were obtained.
Page 98
7.2.3 Influence of the modification on selectivity and reaction rates
The selectivity towards l-bromo-4-aminobenzene is lower if the catalyst is
modified by formaldehyde and higher again, at high modification
concentrations of formaldehyde (Figure 7-8). A systematic prevention of
100-
COCO
CC
E
CDN
CD.Q
O
ÇZ
Ecc
I
<frI
o
Eo
99-
98-
97-
96-
95-
unmodified 1 % CH20
T T
2% CH20 3 5%CH20 5% CH20
Fig. 7-8: Influence of the modification of Raney nickel by formaldehyde on the
hydrogénation of l-bromo-4-nitrobenzene. Reaction conditions: 10 g
l-bromo-4-nitrobenzene, 90 g THF, 60°C. 0.5 MPa and 1500 rpm.
the dehalogenation was not observed. A loss of activity was observed, as it
is demonstrated in Figure 7-9 and Table 7-3.
Page 99
unmodified
• 1%CH20 modified
* 2% CH20 modified
3.5% CH20 modified
5% CH20 modified
0-1,: 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
0 200 400 600 800 1000 1200 1400 1600 1800 2000
time / [min]
Fig. 7-9: Hydrogen uptake (from a 37.5 ml storage vessel thermostated at 25°C) of
modified Raney nickel. Reaction conditions: 10 g l-bromo-4-nitrobenzene,
90 g THF, 60°C. 0.5 MPa and 1500 rpm.
Table 7-3: Summary of the influence of formaldehyde modified Raney catalysts on the
initial hydrogénation rate and the selectivity towards l-bromo-4-
aminobenzene.
modification initial hydrog
rate
enation l-bromo-4-
aminoazobenzene
/[-] / [mmolAV'kg)] / [mass-%)]
unmodified 22.2 99.3
1% CH20 11.5 98.5
2% CH20 7.8 98.4
3.5%CH20 7.5 98.6
5% CH20 4.8 99.4
aniline
/ [mass-%>]
0.7
0.5
1.6
1.4
0.6
Page 100
7.3 Hydrogénation of levodione
7.3.1 General remarks
The low pressure hydrogénation of levodione ((6R)-2,2,6-trimethylcyclo-
hexa-l,4-dione) to actinol ((4R,6R)-4-hydroxy-2,2,6-trimethylcyclohexa-
none) is an important industrial process. Thereby, the R,R isomer is the
desired product [99, 100]. A scheme of the reaction is given in Figure 7-10.
y^° h2 r^V0 r^V0«
+
0A/R\ "H2 HO<T^R% HO* S^^rX
Fig. 7-10: Reaction scheme for the hydrogénation of levodione to R,R- or R,S-actinol.
The reaction was carried out in a 100 ml three-neck sulphonation flask
(Chapter 9.1.3) according the described procedure in Chapter 9.2.5. The
catalyst was washed three times with methanol (Chapter 9.2.8) before the
reaction was started. The reaction conditions were: 10 g levodione, 50 g
MeOH, 25°C, 0.11 MPa and 600 rpm.
7.3.2 Test for a possible gas-liquid transfer limitation for hydrogen
To check a possible influence of a transfer limitation on the hydrogénation
rate, the hydrogen uptake was monitored for two different catalyst amounts
(4 and 6 g Raney nickel). The reaction conditions were: 10 g levodione, 50 g
methanol, 25°C, 0.11 MPa and 600 rpm. The initial hydrogen uptake was
0.071 bar/(min*g) (4 g catalyst) and 0.10 bar/(min*g) (6 g catalyst),
respectively. Thus, a gas-liquid transfer limitation for hydrogen can be
Page 101
excluded, as the rate is linearly proportional to the amount of the catalyst
used.
7.3.3 Influence of the modification on selectivity and hydrogénation
rate
The influence of a modification by formaldehyde on the selectivity towards
the desired isomer and the hydrogénation rate was investigated using
unmodified and differently modified Raney catalysts. The influence on the
hydrogen uptake is plotted in Figure 7-11. The formaldehyde modified
catalysts have less activity then the unmodified. The selectivity towards the
desired R,R-actinol is shown in Figure 7-12. A higher selectivity towards
CD.Q
~ 15-
cu.*:
ro
Q.
cu
o
T3
30 -,
25-
20-
10-
400
time/[min]
Fig. 7-11: Hydrogen uptake during the hydrogénation of levodione using differently
modifiedRaney nickel. Reaction conditions: 10 g levodione, 4 g catalyst, 50 g
MeOH, 25°C. 0.11 MPa and 600 rpm.
Page 102
[CHp]/[%]
Fig. 7-12: Selectivity of the hydrogénation of levodione using differently modified
Raney catalysts. Reaction conditions: 10 g levodione, 4 g catalyst, 50 g
MeOH, 25°C. 0.11 MPa and 600 rpm.
R,R-actinol was not observed. The reported selectivity of 80-85%
R,R-actinol [99] could not be confirmed, even not if fresh catalysts were
used. The distribution of R,R-actinol and R,S-actinol is about 1:1. The
possible isomerisation of the hydrogénation products, can be explained
either by the sodium hydroxide present in the catalyst or by residual acid in
the levodione. Both, acid as well as base, catalyse the isomerisation of the
product.
7.4 Discussion
The hydrogénation of other substances than nitriles was performed to find
other potential applications for formaldehyde treated nickel catalysts. These
systems were screened without a profound study of the reaction mechanism.
Page 103
The modification of nickel catalysts by formaldehyde has no benefit
with respect to a selective hydrogénation of crotonaldehyde. The desired
crotylalkohol is not preferentially produced using the modified catalyst. The
modification only lowers the hydrogénation rates and causes undesired
by-products formed by condensation reactions on the catalyst.
If a halogenated nitrobenzene is hydrogenated, a modification by
formaldehyde does not systematically inhibit the dehalogenation. However,
a decrease in activity was observed.
The hydrogénation of levodione using formaldehyde treated catalysts
does not improve the yield to the desired R,R-actinol. Only a decreased
activity was observed.
Summing up, until now no other hydrogénations than that of nitriles
were found for which a modification of the nickel catalysts by
formaldehyde has been beneficial.
Page 104
Chapter
Conclusions and outlook
The modification of Raney nickel by formaldehyde leads to higher
selectivities towards the primary amines, also in cases that aliphatic nitriles
are hydrogenated. Up to 50% of the formaldehyde used to modify the
catalyst were found as methanol in the modification solution. A possible
explanation for this behavior is that formaldehyde disproportionates at the
catalyst to methanol and chemisorbed carbon monoxide.
If amines have to be produced from nitriles at large scale, several points
have to be thoroughly considered to produce high selectivities:
• The reaction time until the hydrogénation is finished has to be known,
so that the reaction mixture can be cooled immediately after full
conversion (kinetically controlled product distribution) to avoid
disproportionation reactions, that lower the selectivity
(thermodynamically controlled product distribution).
Low hydrogen pressures favour the production of primary amines, thus
a hydrogénation in the gas-liquid transfer limitation for hydrogen (high
catalyst loadings) may be favorable.
The modification of catalysts by formaldehyde offers a new way to
increase the selectivity towards primary amines, especially in cases that
aromatic nitriles are hydrogenated. In order to find the optimal
modification concentration investigations have to be performed because
desired higher selectivities are also accompanied by undesired lower
reaction rates.
An optimal temperature has to be found, because high temperatures lead
to higher selectivities towards primary amines (higher activation energy
of the hydrogénation to the primary amine than of the hydrogénation to
the secondary amine), but disproportionation reactions leading to the
Page 105
8
product distribution determined by the thermodynamic equilibrium can
lower the selectivity at high temperatures [27].
Page 106
Chapter ^
Experimental
9.1 Apparatus
9.1.1 Description of the 500 ml steel hydrogenator
Hydrogénations of butylamine and crotonaldehyde were carried out in a
500 ml hardening vessel (built by F. Hoffmann-La Roche, Figure 9-1). The
apparatus is designed to withstand a maximum pressure of 10 MPa and a
maximal temperature of 200°C. The autoclave is equiped with a four-bladed
agitator that can be rotated at rates between 10 and 2000 rpm.
The reactor can be operated at temperatures between 20 and 200°C. The
heating is electrical and cooling is achieved using cooling water at 15°C.
The autoclave consists of a reaction vessel and a cap, both made of rust-
resistant steel (W. No. 1.4435). A silverseal and 8 screws ensure tightness.
The cap has 7 boreholes (Figure 9-2), and within a central bore for the
agitator shaft (1). Hydrogen is introduced from the side via the cap. Samples
can be taken via valve 4 (Nova Swiss).
Liquid or gaseous starting products are added via valve 7 (Nova Swiss).
The temperature is measured with a type PT100 sensor (range: -100 to
400°C, precision: ±0.33%) manufactured by Rotax. This sensor is
positioned in borehole 2. The temperature is controlled by cascade
regulation.
The pressure in the autoclave is measured with a piezometer PA-23 100
(range: 0.1 to 10 MPa, precision: ±0.05 MPa) manufactured by Rotax (8).
The pressure in the hydrogen storage vessel is measured using a piezometer
PA-23 200 (range: 0.1 to 20 MPa, precision ±0.1 MPa) also manufactured
by Rotax. A bursting disk (Sitec) is also installed (bursting pressure: 11 MPa
Page 107
degassing
sampling system
temperature measurement
V\/\/WVWV\/W\^ / /W\A NW\MVW\AVWWWWWWWVWWWW\
copper block with electrical heating and cooling
hydrogen storage vessel
09
Fig. 9-1 : Hydrogénation apparatus.
±10%) in the cap (9). The autoclave pressure can be released by a valve (3,
Sitec). Nitrogen is fed to the autoclave via valve 6 (Nova Swiss). The
remaining holes (5 and 7) were not used and were therefore fitted with blind
flanges.
Page 108
1 agitator shaft
2 temperature measurement
3 pressure release
4 valve for sampling5 blind flange6 nitrogen inlet
7 blind flange8 pressure measurement
9 bursting disk
Fig. 9-2: Hydrogénation apparatus: Top view of the cap.
9.1.2 200 ml glass hydrogenator
A 200 ml glass hydrogénation vessel from Bilchi AG (max. pressure 1 MPa,
max. temperature 200°Cj was mounted on an apparatus containing a Bilchi
Cyclo 075 stirrer (max. speed 3000 rpm), a PA-23 10 piezometer (Rotax,
range 0.1 to 1 MPa), a PT 100 B (Rotax, -100 to 400°C, ±0.33%) and a
bursting disk (Sitek, 1 MPa bursting pressure). Two hydrogen storages,
thermostated at 25°C with a Lauda E 100 thermostat, were used (5 ml and
50 ml, total hydrogen storage volume inclusive pipes: 73.5 ml) during the
reaction. The glass hydrogenator was thermostated with a Lauda E 200
thermostate. The reaction data were stored in aEurotherm Chessel 4100 G
9.1.3 100 ml low pressure hydrogénation apparatus
A three-neck sulphonation flask, equiped with a glass thermometer and a
gas inlet (inlet of argon, hydrogen and pressure release on the same pipe)
was mounted on a stirrer system (the third hole was fitted with a blind
flange). A Julabo PC/4 thermostat was used to control the reaction
Page 109
temperature. The hydrogen uptake was monitored by a W+W recorder 312
from Kontron AG attached to a piezometer in the storage vessel. A bursting
disk (Sitek) was installed with a bursting pressure of 0.07 MPa.
9.1.4 Lab Shaker
A Lab Shaker from KühnerAG with 35 ml hydrogénation reactors (built by
F. Hoffmann-La Roche) was used to perform the reversibility experiments.
These autoclaves were equiped with a bursting disk (30 MPa) and a valve.
A shaking frequency of 280 min"1 on 25 mm was applied.
9.1.5 Modification and washing apparatus
Modification and washing procedures were carried out in a three-neck
sulphonation flask, equiped with a nitrogen intake and an outlet connected
to a bubble counter. A glass stirrer driven by a motor was installed in the
central grinding. If the catalyst was modified, a dropping funnel was
installed in the last vent, otherwise this vent was provided with a blind plug.
9.1.6 Gas Chromatograph
An HP 6890 Series GC System with an HP 7683 Series Injector
autosampler was used. A flame ionisation detector was used for normal
measurements. An HP 5973 Mass Selective Detector was used for gas
chromatography with mass spectroscopic detection. HP ChemStation
software, version Rev.A. 06.03 [509] was used to control and analyse the
measurements. By aid of this program an analysis method was developed
[101-103].
Page 110
9.2 Methods
9.2.1 Hydrogénation of butyronitrile
240 g (3 mol) butyronitrile were placed in the 500 ml steel autoclave
(Chapter 9.1.1) before the catalyst (if Raney nickel was used, the catalyst
was added as wet paste) was added. The autoclave was sealed and purged
three times with 0.55 MPa nitrogen. The reactor was then heated while the
content was stirred slowly (600 rpm). Once the reaction temperature had
been reached, the stirrer was turned off and hydrogen was pressed into the
autoclave. The reaction was initiated by turning on the stirrer. Samples were
taken during the hydrogénation (Chapter 9.2.6).
After the reaction was completed, the reaction vessel was cooled to
room temperature, the pressure released and the vessel purged three times
with 0.55 MPa nitrogen. The reaction mixture was sucked off by a vacuum
pump and then filtered. The autoclave and the filtering apparatus were
purged with methanol. To clean the autoclave, the vessel was filled with
methanol which then was boiled at 80°C for 15 min.
The Raney nickel was disposed into a special nickel waste jar and the
reaction mixture in a solvent waste jar.
9.2.2 Reversibility experiments in the 35 ml screening autoclave
20 g butylamine (0.27 mol) were placed in a 35 ml screening hydrogenator
before 2 g of wet Raney nickel catalyst were added. The hydrogenator was
then closed and purged three times with 4.2 barg nitrogen. Ammonia or
hydrogen were then added via the valve on the autoclave. The autoclave was
placed in the Lab Shaker and heated to reaction temperature while being
shaken at 280 min"1.
After a defined reaction time the screening hydrogenator was placed in
a cooling block (copper block cooled by water), the pressure was released,
the reactors were again purged three times with 4.2 barg nitrogen and then
Page ill
opened. The content of the reactors was flushed out with methanol. The
reaction mixture was then filtered and analysed.
9.2.3 Hydrogénation of crotonaldehyde
40 g (0.57 mol) butyronitrile and 200 g ethanol were placed in the 500 ml
steel autoclave (Chapter 9.1.1) before the catalyst (ifRaney nickel was used,
the catalyst was added as wet paste) was added. The autoclave was sealed
and purged three times with 0.55 MPa nitrogen. The reactor was then heated
while the content was stirred slowly (600 rpm). Once the reaction
temperature had been reached, the stirrer was turned off and hydrogen was
pressed into the autoclave. The reaction was initiated by turning on the
stirrer. Samples were taken during the hydrogénation (Chapter 9.2.6).
After completion of the reaction, the reaction vessel was cooled to room
temperature, the pressure released and the vessel purged three times with
0.55 MPa nitrogen. The reaction mixture was sucked off with a vacuum
pump and then filtered. The autoclave and the filtering apparatus were
purged with ethanol. To clean the autoclave, the vessel was filled with
ethanol which then was boiled at 80 °C for 15 min.
The Raney nickel was disposed into a special nickel waste jar and the
reaction mixture in a solvent waste jar.
9.2.4 Hydrogénation of l-bromo-4-nitrobenzene
10 g (50 mmol) l-bromo-4-nitrobenzene and 100 g tetrahydrofurane were
placed in the 200 ml glass autoclave (Chapter 9.1.2) before the catalyst was
added as a wet paste. The autoclave was sealed and purged three times with
0.3 MPa nitrogen. The reactor was then heated while the contents were
stirred slowly (600 rpm). Once the reaction temperature had been reached,
the stirrer was turned off and hydrogen was pressed into the autoclave. The
reaction was initiated by turning on the stirrer.
After completion of the reaction, the reaction vessel was cooled to room
temperature, the pressure released and the vessel purged three times with
Page 112
0.3 MPa nitrogen. The reaction mixture was sucked off with a vacuum
pump and then filtered. The autoclave and the filtering apparatus were
finally purged with tetrahydrofurane.
The Raney nickel was disposed into a special nickel waste jar and the
reaction mixture in a solvent waste jar.
9.2.5 Hydrogénation of levodione
10 g (65 mmol) levodione ((6R)-2,2,6-trimethylcyclohexa-l,4-dione) and
50 g methanol were placed in a 100 ml three-neck sulphonation flask
(Chapter 9.1.3) before the catalyst was added as a wet paste. The flask was
installed on the hydrogénation equipment, evacuated and purged three times
with 0.11 MPa argon. The reactor was then heated while the contents were
stirred slowly (600 rpm). Once the reaction temperature had been reached,
the stirrer was turned off and hydrogen was pressed into the autoclave.
Then, the reaction was started by turning the stirrer on again.
After the reaction was completed, the reaction vessel was cooled to
room temperature, the pressure released and the vessel evacuated and
purged three times with 0.11 MPa argon. The reaction mixture was sucked
off with a vacuum pump and then filtered. To clean the autoclave and the fil¬
tering apparatus they were purged with methanol.
The Raney nickel was disposed into a special nickel waste jar and the
reaction mixture in a solvent waste jar.
9.2.6 Description of the sampling procedure
Samples were taken with a steel capillary with 1 mm inner diameter and a
2 jim frit (dead volume: 0.5 ml). Because sometimes the reaction
temperatures were higher than the boiling temperatures of the substances,
the samples were collected in a test tube with solvent (methanol or ethanol in
the case of butyronitrile or crotonaldehyde, respectively). These test tubes
were cooled in a Dewar vessel containing dry ice and ethanol. Initially, 2 ml
ofthe reaction mixture were rejected before 1 ml was collected and analysed.
Page 113
9.2.7 Neutralisation with water
50 g Raney nickel were suspended in 100 ml distilled water in a three-neck
sulphonation flask. The flask was purged three times with argon before the
suspension was stirred for 30 min. The catalyst was then allowed to settle
before the solvent was decanted. This procedure was repeated three times
with distilled water. The neutralised catalyst was stored in distilled water.
9.2.8 Neutralisation with methanol
50 g Raney nickel were suspended in 100 ml distilled water in a three-neck
sulphonation flask. The flask was purged three times with argon before the
suspension was stirred for 30 min. The catalyst was then allowed to settle
before the solvent was decanted. This procedure was first carried out once
with distilled water and then three times with methanol. The neutralised
catalyst was stored in methanol.
9.2.9 Neutralisation with tetrahydrofurane
50 g Raney nickel were suspended in 100 ml distilled water in a three-neck
sulphonation flask. The flask was purged three times with argon before the
suspension was stirred for 30 min. The catalyst was then allowed to settle
before the solvent was decanted. This procedure was first carried out once
with distilled water and then three times with tetrahydrofurane. The
neutralised catalyst was stored in tetrahydrofurane.
9.2.10 Modification with formaldehyde
50 g Raney nickel were suspended in 100 ml distilled water in a three-neck
sulphonation flask. The flask was purged three times with argon before
13 ml of a solution of 35% formaldehyde were added slowly. The
suspension was stirred for 30 min. The catalyst was then allowed to settle
before the solvent was decanted. The catalyst was washed three times with
Page 114
30 ml distilled water (in accordance with Chapter 9.2.7), and the modified
catalyst was stored in methanol.
9.3 Analytics
9.3.1 Determination of butyronitrile, butylamine, dibutylamine and
dibutylimine with a GC method using an internal standard
GC conditions
apparatus HP 6890 gas Chromatograph with
split injector and FID
HP 7863 autosampler
column stationary phase Rtx-5 Amine
length x ID 30 m x 0.32 mm.,
film 1.0 jim
column material 5% diphenyl- 95% dimethyl polysiloxane
manufacturer Restek Corporati;on
carrier gas helium pressure 75kPa
total flow 112 ml/min
split ratio 50:1
column temperature 65°C(CImin), 3°/min, 80°C(0min), 20°C/min;
280°C(0 min)
injector temperature 250°C
detector temperature 250°C
injection volume ljLLl
Sample preparation
ISTD solution 5 gn-caprylic acid methyl ester were diluted in 1 ]
methanol.
Page 115
calibration solution
sample solution
5-10 mg of reference substances were weighted in
a sample vial before 1 ml internal standard
solution was added.
15 jil of the collected sample were diluted in 1 ml
internal standard solution.
analysis time
retention times butylamine
butyronitrile
dibutylimine
dibutylamine
ISTD
tributylamine
15.0 min
2.5 min
2.9 min
7.4 min
8.0 min
10.1 min
10.6 min
Remarks
Calibrations were performed using four different concentrations of the
substances to be analysed.
9.3.2 Determination of crotonaldehyde, crotylalkohol, butanal and
butanol with a GC method using an internal standard
GC conditions
apparatus HP 6890 gas Chromatograph with
split injector and FID
HP 7863 autosampler
column stationary phase
length x ID
column material
manufacturer
Stabilwax
30 m x 0.32 mm, film 0.25 jim
Carbowax-PEG
Restek Corporation
Page 116
carrier gas helium pressure 75kPa
total flow 127 ml/min
split ratio 50:1
column temperature 40°C(5min), 37min,
240°C(3.6min)
70°C(0min), 15°C/min
injector temperature 240°C
detector temperature 240°C
injection volume l|il
Sample preparation
ISTD solution 5 g n-caprylic acid metlîyl ester were diluted in 1
calibration solution
sample solution
ethanol.
5-10 mg of reference substances were weighted in
a sample vial before 1 ml internal standard
solution was added.
100 jil were diluted in 1 ml internal standard
solution.
analysis time
retention times butanal
crotonaldehyde E
crotonaldehyde Z
butanol
crotylalkohol E
crotylalkohol Z
ISTD
30.0 min
2.4 min
4.9 min
5.0 min
8.5 min
11.7 min
12.9 min
17.8 min
Remarks
Crotonaldehyde and crotylalkohol were calibrated and measured as the sum
of the E and Z isomers. Four samples of different concentrations were used
for calibration.
Page 117
9.3.3 Determination of l-bromo-4-nitrobenzene, l-bromo-4-
aminobenzene and aniline
GC conditions
apparatus HP 6890 gas Chromatograph with
split injector and FID
HP 7863 autosampler
column stationary phase DB-5HT
length x ID 15 m x 0.25 mm, film 0.1 jim
column material 5% diphenyl - 95% dimethyl polysiloxane
manufacturer J&WScientific
carrier gas helium pressure 57 kPa
total flow 33 ml/min
split ratio 25:1
column temperature 60°C(0min), 107min, 250°C(0min)
injector temperature 250°C
detector temperature 250°C
injection volume l|il
Sample preparation
sample solution 500 jil of the filtrated reaction mixture wen
diluted with 500 jllI THF.
analysis time 19.0 min
retention times aniline 1.9 min
1 -bromo-4-nitrobenzene
1 -bromo-4-aminobenzene
2.9 min
5.2 min
Page 118
9.3.4 Determination of levodione and actinol
GC conditions
apparatus HP 6890 gas Chromatograph with
split injector and FID
HP 7863 autosampler
column stationary phase
length x ID
column material
manufacturer
HP-5
30 m x 0.32 mm, film 0.25 jim
5% diphenyl - 95% dimethyl polysiloxane
Hewlett Packard
carrier gas helium pressure
total flow
split ratio
79 ml/min
50:1
column temperature
injector temperature
detector temperature
injection volume
50°C(0min), 107min, 150°C(3min)
250°C
300°C
ljLLl
Sample preparation
sample solution 100 jil of the filtrated reaction solution were
diluted with 900 jil methanol
analysis time
retention times levodione
(R,S)-actinol
(R,R)-actinol
13.0 min
8.8 min
9.8 min
9.9 min
Page 119
9.3.5 Methanol determination in aqueous medium with a headspace GC
method using an external standard
GC conditions
apparatus Agilent 6890N gas Chromatograph with FID
PE HS 40 XL headspace autosampler
column stationary phase
length x ID
column material
manufacturer
DB-1
30 m x 0.32 mm, film 5 jim
100% dimethyl polysiloxane
J&WScientific
carrier gas nitrogen pressure 20 ps
splitless
column temperature 40°C(3min), 107min, 150°C(2min)
transfer temperature 100°C
injector temperature 140°C
detector temperature 300°C
injection time 0.08s
Sample preparation
calibration solution
sample solution
a solution of 100 jil MeOH in 1000 ml water was
prepared as ESTD, 1 ml of this solution was
placed in the HS autosampler.
the aqueous modification solution was diluted
1:10 with water before 100 mg were again diluted
in 1 ml water and then placed in the HS
autosampler.
analysis time
retention times methanol
16.0 min
1.6 min
Page 120
9.3.6 Formaldehyde determination in aqueous medium with an HPLC
method using an external standard
HPLC conditions
apparatus HP 1050 HPLC with UV detector
column stationary phase
length x ID
manufacturer
Supelcosil LC-18
250 mm x 4.6 mm, film 5 jim
Supelco
solvent acetonitrile : water 60 : 40
flow 1.5ml/min
detection wavelength
injection volume
360 nm
2 |il
Sample preparation
derivatization solution
derivatization
calibration solution
2.37 g (8 mmol) 2,4-dinitrophenylhydrazine were
dissolved in 100 ml THF.
5 ml of the derivatization solution were tared,
before 30 mg of the sample were added. 1 drop of
cone, hydrochloric acid was added and the sample
was stored for 1 hour at 50°C. 1 ml of the warm
derivatized solution was diluted with 10 ml
acetonitrile and then analysed.
a solution containing 5% formaldehyde was
derivatized and then used as ESTD.
analysis time
retention times 2,4-dinitrophenylhydrazone
20.0 min
3.8 min
Page 121
9.3.7 Synthesis of dibutylimine as a standard for GC measurements
Butylamine condenses with butyraldehyde to produce the Schiff base
(Figure 9-3) [104, 105].
Fig. 9-3: Condensation of butylamine and butyraldehyde [104].
34.6 g (506 mmol) butyraldehyde in 56 g toluene were placed in a three-
neck sulphonation flask before 36.8 g (504 mmol) butylamine were added
slowly via a dropping funnel. The solution was heated to 66°C and water
was produced. After one hour the aqueous phase was separated off and the
organic phase was distilled.
The same conversion was also made on a smaller scale (calibration
samples for gas chromatographic analysis) and on a larger scale (to produce
azomethine as starting material for the hydrogénation), in both cases
without solvent and using an excess of butylamine. This conversion is
quantitative, no butyraldehyde was detected after the reaction.
The condensation of dibutylamine and butyraldehyde is slower than the
one of butylamine and butyraldehyde, but using the same conversion,
tributyleneamine could be produced. This was not done because
tributyleneamine was not detected during the hydrogénation of butyronitrile
in the reaction mixture.
Page 122
9.4 Identification of by-products
9.4.17V-Butylbutanamide
Initially, a reaction mixture, containing 4 mass-% TV-butylbutanamide was
concentrated and distilled under vacuum. The obtained sample was purified
by chromatography using ethyl acetate/light petroleum (50 : 50) as eluent to
yield the pure compound. A DPX 400 NMR spectrometer (400 MHz 1H, 75
MHz 13C) was used to measure the nuclear magnetic resonance spectra;
Ôh(CDC13) 0.92 (3H, t, .7=8.0, CH3), 0.95 (3H, t, .7=8.2, CH3), 1.34 (2H, m,
C772CH2CH2NH), 1.48 (2H, m, C772CH2NH), 1.66 (2H, sext, J=l .5
C772CH2CO), 2.15 (2H, t, .7=7.4, CH2CO), 3.25 (2H, q, .7=6.7, CH2NH),
5.67 (1H, s, NH); 5C(CDC13) 14.1 (2CH3), 19.6 (CH2CH2CO), 20.4
(CH2CH2CH2N), 32.1 (CH2CH2N), 39.2 (CH2CO), 39.6 (CH2N) and 173.4
(CO). A mass spectrum was measured using a GC-MS system (5973 Mass
Selective Detector); m/z 143 (M+, 19%), 128 (C7H14NO+, 23), 115 (29), 101
(25), 100 (C5H10NO+, 57), 88 (20), 86 (C4H8NO+, 19), 73 (52), 71 (C4H70+,
90), 57 (C4H9+, 35), 44 (62), 43 (C3H7+, 95), 41 (60) and 30 (C2H6+, 100).
These results are in accordance with published data [106].
9.4.2 7V,7V-Dibutylbutyramidine
The reaction mixture containing about 7 mass-% AyV-dibutylbutyramidine
was concentrated to a small volume and then distilled in vacuo. A sample of
15 mass-% TV-butylbutanamide and 85 mass-% AyV-dibutylbutyramidine
was obtained. An AV 500 NMR spectrometer (500 MHz 1H, 125 MHz 13C,36 MHz 14N (ref. nitromethane)) was used to measure the nuclear magnetic
resonance spectra; 5H(CDC13) 0.94 (6H, t, .7=7.4, CH3CH2CH2CH2N), 1.09
(3H, t, .7=7.4 CH3CH2CH2C), 1.37 (4H, m, C772CH2CH2N), 1.63 (6H, m,
C772CH2N, C772CH2C), 2.34 (2H, m, CH2C), 3.19 (4H, t, .7=7.5 CH2N),
7.29 (1H, s, NH); 5C(CDC13) 13.6 (2 CH3CH2CH2CH2N), 14.1
(CH3CH2CH2C) 19.7 (2 CH2CH2CH2N), 20.0 (2 CH2CH2N), 27.5
(CH2CH2C), 31.8 (CH2C), 43.3 (2 CH2N), 137.6 (CNH); ôN(CDCl3, 300K)
Page 123
-165, line width 90, -280, line width 75; ÔN(CDC13, 333K) -230 line width
175; ôN(TFA/dioxane, 300K) -260, line width 125. An NMR literature
reference for aromatic amidines is given in [107]. A mass spectrum was
measured using a GC-MS system (5973 Mass Selective Detector); m/z 198
(M+, 20%), 183 (CnH23N2+, 24), 169 (C10H21N2+, 29), 155 (C9H19N2+, 17),
141 (C8H17N2+, 17), 127 (C8H17N+, 45), 113 (16), 99 (15), 85 (15), 84 (78),
72 (C4H10N+, 26), 70 (C4H8N+, 100), 57 (C4H9+, 17), 43 (C3H7+, 24), 41
(33)and29(C2H5+,23).
9.4.3 1,1-Diethoxybutane
A mass spectrum was measured using a GC-MS system (5973 Mass
Selective Detector); m/z 103 (C5Hn02+, 100%), 101 (C6H130+, 85), 75 (50),
73 (M2+, 40), 55 (48), 47 (38), 43 (C3H70+, 16) and 29 (C2H50+, 14).
9.5 Characterisation of the catalyst
9.5.1 The reduction potential
The reduction potential of Raney nickel was measured with two different
combined gold electrodes with an integrated Ag/AgCl reference, model
6.0413.100, manufactured by Metrohm. The measurement was monitored
by a Metrohm pH-meter 691. The electrodes were calibrated with Metrohm
redox standard 6.2306.020 (U=250±5 mV). When the redox potential was
measured, the electrodes were placed directly on the solid catalyst. The
values obtained were converted into the standard reduction potentials
relative to the Pt/H2-electrode.
Page 124
9.5.2 Adsorption of an indicator
0.01-0.02 g of wet Raney nickel were placed in a 100 ml volumetric flask
before 100 ml of a 10"4 M 4-aminoazobenzene solution were added. The
solution was allowed to stand for 4 weeks, before the concentration of
4-aminoazobenzene was measured with an UVIKON 720 LC UV-VIS
absorption spectrometer. From the difference in concentration the amount of
the adsorbed indicator was calculated.
9.5.3 Dissolution in acidic medium
1 g of wet Raney nickel was placed in a three-neck sulphonation flask
before 50 g water were added. With a dropping funnel 32 ml of hydrochloric
acid (35%) were slowly added while the content of the flask was stined with
a magnetic stirrer. The released gas was measured with a measuring cylinder
that was filled with water inversely placed in a water containing vessel.
9.6 Chemicals
The chemicals used, their suppliers and purity grades are listed in Table 9-1.
Table 9-1 : Purity values and suppliers of the chemicals used
Substance Supplier Purity grade
4-aminoazobenzene Merck for synthesis > 98% (HC104)
aniline Fluka puriss. p. a. > 99.5% (GC)
l,2-bis(2-hydroxyethyl- F. Hoffmann-La Roche Lot No. 28611
thio)ethane (EDS)
1 -bromo-4-nitrobenzene Fluka puriss. p. a. > 98% (GC)
butanol Fluka puriss. p. a. > 99.5% (GC)
Page 125
Table 9-1 : Purity values and suppliers of the chemicals used
Substance Supplier Purity grade
butyraldehyde Fluka purum > 97% (GC)
butylamine Fluka purum > 98% (GC)
butyronitrile Fluka purum > 99% (GC)
n-caprylic acid methyl ester Fluka puriss > 99% (GC)
crotonaldehyde (E + Z) Fluka purum > 98% (GC)
crotylalkohol (E + Z) Fluka purum > 97% (GC)
dibutylamine Fluka puriss. > 99% (GC)
ethanol Merck p. a.
formaldehyde solution Merck extra pure 35%
formaldehyde solution Fluka puriss. p. a. ACS free of acids
hydrochloric acid 25% Merck GR for analysis
levodione F. Hoffmann-La Roche Lot No. 32716-a8
methanol Merck p. a.
methanol F. Hoffmann-La Roche tech.
nickel-on-carrier Engelhard Ni 1404 P, Lot H-99
pyridine Fluka p. a. > 99.8% (GC)
tetrahydrofurane Fluka puriss. p. a. > 99.5% (GC)
toluene Fluka puriss. p. a. > 99.5% (GC)
tributylamine Fluka puriss. p. a. > 99% (GC)
argon Carbagas 99.995%
nitrogen Carbagas 99.995%
hydrogen Carbagas 99.995%
Raney nickel Degussa-Hülls AG Type B113 Z, Batch 20018989
Page 126
9.7 Calculation of selectivity and reaction rates
In every experiment samples were taken after 10, 20, 30, 40, 50, 60, 75,
90, 105, 120, 150, 180 min and, if necessary, every further hour. With aid of
these samples reaction profiles were drawn (an example is given in
Figure 9-4) and the selectivity and the initial reaction rates were calculated.
100-
„80-
t/5to
ce60-
's 4°-
CDÜ
OÜ
20-
• m
— butylamine
• butyronitrile* dibutylimine
— dibutylamine
Br em.__
gfTTTTV
-
0 50 100 150 200 250 300 350 400
time / [min]
Fig. 9-4: Typical reaction profile for the hydrogénation of butyronitrile. Reaction
conditions: 240 g BN, 3.75 g RaNi, 100°C, 1 MPa overall pressure,
1000 rpm.
The selectivities of the substances were calculated with the following
equation (Eq. 9.1):
selectivity(x) = mass of component x / total mass Eq. 9.1
Page 127
Initially, to calculate the initial reaction rates, polynomials of second order
were fitted to the data in a range of 0 to 50% butyronitrile conversion. The
reaction rates were then obtained from the first derivatisation of the fitted
function at the point of zero butyronitrile conversion (origin). The obtained
values /[mass-%/min] were converted to molar rates, and these were divided
by the mass of charged catalyst in order to obtain the molar rates per mass
catalyst and second (/[mmol/kg*s]).
9.8 Error analysis
9.8.1 Precision of a gas chromatographic analysis
The precision of an analysis by gas chromatography was measured with a
calibration sample. Five measurements were carried out (Table 9-2). The
relative standard deviation was 1.2% in the worst case.
Table 9-2: Standard deviation (g) and relative standard deviation (rel. g) of a gas chro¬
matographic measurement.
substance
butylamine
butyronitrile
dibutylamine
tributylamine
mean G rel. G
/ [mass-%)] /[mass-%)] /[%]
18.55 0.15 0.81
40.48 0.44 1.09
30.17 0.36 1.19
10.81 0.09 0.84
Page 128
9.8.2 Precision of a sample analysed with gas chromatography
64.14 0.27 0.42
20.87 0.49 2.33
6.52 0.14 2.14
8.46 0.21 2.52
The precision of the sampling procedure with gas chromatography was
measured in experiment 5 (conditions: 240 g BN, 15g methanol washed
RaNi, 100°C, 1 MPa, 1000 rpm) after a reaction time of 120 min. Five
samples were taken and analysed (Table 9-3).
Table 9-3: Standard deviation (g) and relative standard deviation (rel. g) of different
samples.
substances mean g rel. g
/ [mass-%] / [mass-%] / [%]
butylamine
butyronitrile
dibutylamine
dibutylimine
The relative standard deviation was 2.6% in the worst case. It must be noted
that the sampling procedure for five samples lasted 3 min and that during
this time the reaction was running. This value is in an expected range and is
comparable to literature values of about 2% [27].
Concerning the accuracy of the measurement the following has to be
mentioned: To calculate the relative concentrations, all compounds in the
sample were determined in [mg] and divided by the total mass. Compounds
not calibrated were considered using a relative response factor of 1.0
relative to the internal standard. A systematic enor is therefore made when
calculating the total mass and of course in the concentrations of the
components. This enor depends on the concentration of substances that are
not calibrated. Two not calibrated compounds were detected: dibutylamide
and tributylamidine.
Page 129
9.8.3 Precision of the selectivity and the reaction rates
The precision of the selectivity and the reaction rates was measured by
repeating an experiment twice (Table 9-4 and Table 9-5). Different
calibrations of the gas Chromatograph were used and the time periods
between the experiments were several months. The conditions were: 240 g
BN, 15 g RaNi, 100°C, 1 MPa hydrogen, 1000 rpm.
Table 9-4: Precision of the selectivity of the hydrogénation of butyronitrile. Reaction
conditions: 240 g butyronitrile, 15 g RaNi, 100°C, 1 MPa, and 1000 rpm.
BA DBA Tamidine max. DBI
/ [mass-%)] / [mass-%)] / [mass-%)] / [mass-%)]
exp 1 90.98 8.50 0.53 10.80
exp 2 90.73 8.84 0.43 9.73
exp 3 91.13 8.07 0.61 10.46
mean 90.95 8.47 0.52 10.33
standard deviation 0.20 0.38 0.09 0.55
Table 9-5 : Precision ofthe initial reaction rates. Reaction conditions: 240 g butyronitrile,
15 g RaNi, 100°C, 1 MPa, and 1000 rpm.
d[BN]/dt d[BA]/dt d[DBA]/dt d[DBI]/dt
/ [mmol/kg*s] / [mmol/kg*s] / [mmol/kg*s] / [mmol/kg*s]
exp 1 65.7 46.6 1.40 7.97
exp 2 60.7 43.6 1.16 7.35
exp 3 67.0 48.7 0.97 8.17
mean 64.5 46.3 1.18 7.83
standard deviation 3.3 2.5 0.21 0.43
Page 130
Chapter
Literature
[I] O. G Degischer, F. Roessler, US Patent 2001004672, 2001.
[2] S. Nishimura, in "Handbook of Heterogeneous Catalytic
Hydrogénation for Organic Synthesis", John Wiley & Sons, New
York, 2001, pp. 254-285.
[3] P. N. Rylander, in "Catalytic Hydrogénation in Organic Synthesis",
Academic Press, New York, 1979, pp. 138-152.
[4] P. N. Rylander, in "Hydrogénation Methods", Academic Press, New
York, 1985, pp. 94-103.
[5] E. Müller, O. Bayer, in "Methoden der organischen Chemie",
Thieme Verlag, Stuttgard, 1980, IV/lc, Reduktion Teil 1, pp.
110-144.
[6] F. Zymalkowski, in "Katalytische Hydrierungen im Organisch-
Chemischen Laboratorium", Enke Verlag, Stuttgart, 1965, pp.
237-257.
[7] R. Kuhn, W. Kirschenlohr,Liebigs Ann. Chem., 1956, 600, 115-125.
[8] R. Kuhn, H. Grassner, Liebigs Ann. Chem., 1958, 612, 55-64.
[9] E. Möltgen, P Tinapp, Liebigs Ann. Chem., 1979, 1952-1959.
[10] J. Braun, G Blessing, F. Zobel, Ber, 1923, 36, 1988-2001.
[II] H. Greenfield, Ind. Eng. Chem. Prod. Res. Dev, 1967, 6, 142-144.
[12] J. Volf, J. Pasek, Stud. Surf. Sei. Catal., 1986, 27, 105-144.
[13] J. L. Dallons, A. van Gysel, G. Jannes, Chem. Ind. (Dekker), 1992,
47,93-104.
[14] M. J. F. M. Verhaak, A. J. van Dillen, J. W. Geus, J. Catal., 1993,
143, 187-200.
[15] M. J. F. M. Verhaak, A. J. van Dillen, J. W. Geus, Appl. Catal. A,
1993,705,251-269.
Page 131
[16] M. J. F. M. Verhaak, A. J. van Dillen, J. W. Geus, Catal. Lett., 1994,
26,37-53.
[17] Y. Y. Huang, W. M. H. Sachtler, J. Phys. Chem. B, 1998, 102,
6558-6565.
[18] Y. Y. Huang, W. M. H. Sachtler, Appl. Catal. A, 1999,182, 365-378.
[19] Y. Y. Huang, W. M. H. Sachtler, J. Catal., 1999,184, 247-261.
[20] Y. Y. Huang, W. M. H. Sachtler, J. Catal., 1999, 7 88, 215-225.
[21] Y. Y. Huang, W. M. H. Sachtler, J. Catal, 2000,190, 69-1A.
[22] Y. Y. Huang, W. M. H. Sachtler, Stud. Surf. Sei. Catal, 2000,130 A,
527-532.
[23] C. V Rode, M. Arai, M. Shirai, Y. Nishiyama, Appl. Catal. A, 1997,
148, 405-413.
[24] B. Coq, D. Ticht, S. Ribet, J. Catal, 2000,189, 117-128.
[25] M. Besson, J. M. Bonnier, M. Joucla, D. Djaouadi, Bull. Soc. Chim.
Fr, 1990, 727, 5-19.
[26] C. de Bellefon, P. Fouilloux, Catal. Rev Sei. Eng, 1994, 36,
459-506.
[27] O. G. Degischer, Dissertation Nr. 14012, ETH Zürich, 2001.
[28] F. M. Cabello, D. Tichit, B. Coq, A. Vaccari, N. T. Dung, J. Catal,
1997, 7<57, 142-152.
[29] H. R. Christen, F. Vögtle, in "Organische Chemie, von den
Grundlagen zur Forschung", Otto Salle Verlag, Frankfurt, 1992, 1,
pp. 561-830.
[30] S. N. Thomas-Pryor, T. A. Manz, Z. Liu, T. A. Koch, S. K.
Sengupta, W. N. Delgass, Chem. Ind. (Dekker), 1998, 75, 195-206.
[31] T. A. Johnson, US Patent 5869653, 1999.
[32] M. Besson, D. Djaouadi, J. M. Bonnier, S. Hamar-Thibault, M.
Joucla, Stud. Surf Sei. Catal, 1991, 59, 113-120.
[33] O. Nicodemus, W. Schmidt, DE Patent 510439, 1930.
[34] P. Herold, K. Smeykal, DE Patent 626923, 1936.
[35] W. Reppe, E. Bauer, DE Patent 741683, 1943.
[36] H. Raab, DE Patent 738448, 1943.
[37] E. Hädicke, W. Schlenk, Liebigs Ann. Chem, 1972, 764, 103-111.
Page 132
[38] A. Baiker, W Caprez, W L. Holstein, Ind. Eng. Chem. Prod. Res.
Dev, 1983, 22, 217-225.
[39] J. F. Olin, T. E. Deger, US Patent 2192523, 1940.
[40] S. Nishimoto, B. Ohtani, T. Yoshikawa, T. Kagiya, J. Am. Chem.
Soc,1983, 705,7180-7182.
[41] M. J. F. M. Verhaak, A. J. van Dillen, J. W. Geus, Appl. Catal. A,
1994,109, 263-275.
[42] P. Sabatier, in "Die Katalyse in der Organischen Chemie",
Akademische Verlagsgemeinschaft, Leipzig, 1927.
[43] M. Raney, US Patent 1563587, 1925.
[44] M. Raney, US Patent 1628190, 1927.
[45] P Fouilloux, Appl. Catal, 1983, 8, 1-42.
[46] S. Nishimura, in "Handbook of Heterogeneous Catalytic
Hydrogénation for Organic Synthesis", John Wiley & Sons, New
York, 2001, pp. 7-19.
[47] K. Weissermel, H. J. Arpe, in "Industrial Organic Chemistry", VCH,
1993, pp. 64-72.
[48] G Ertl, H. Knözinger, J. Weitkamp, in "Handbook of Heterogeneous
Catalysis", VCH, 1997.
[49] R. L. Augustine, in "Heterogeneous Catalysis for the Synthetic
Chemist", Dekker, New York, 1996, pp. 473-510.
[50] P. Marion, US Patent 2001027257, 2001.
[51] G A. Martin, P. Fouilloux, J. Catal, 1975, 38, 231-237.
[52] D. Djaouadi, M. Besson, J. Jenck, P. Fouilloux, Chem. Ind.
(Dekker), 1995, 62, 423-429.
[53] J. R. Kosak, US Patent 3546297, 1970.
[54] J. F. Mais, K. C. Paetz, H. Fiege, H. U. Blank, D. Brueck, W. Mehl,
DE Patent 19604988, 1998.
[55] J. L. Dallons, G Jannes, B. Delmon, Stud. Surf. Sei. Catal, 1988, 41,
115-121.
[56] H. Lei, Z. Song, D. Tan, X. Bao, X. Mu, B. Zong, E. Min, Appl.
Catal. A, 2001, 214, 69-76.
[57] G Wegener, E. Waldau, B. Pennemann, B. Temme, H. Warlimont,
U. Kühn, DE Patent 19753501, 1999.
Page 133
[58] S. D. Mikhailenko, T. A. Khodareva, E. V. Leongardt,
A. I. Lyashenko, A. B. Fasman,J. Catal, 1993,141, 688-699.
[59] J. R. Kosak, US Patent 3145231, 1964.
[60] GB Patent 919273, 1963.
[61] P. Baumeister, W. Schener, US Patent 4960936, 1990.
[62] M. Studer, P. Baumeister, US Patent 6096924, 2000.
[63] O. G. Degischer, Laboruntersuchung zur Hydrierung von Pynitril,
VFCR, F. Hoffmann-La Roche, 2000.
[64] R. C. Balk, Benzonitril Hydrierung, VFCR, F. Hoffmann-La Roche,
2000.
[65] D. E. Gardin, G A. Somorjai, J. Phys. Chem., 1992, 96, 9424-9431.
[66] P D. Ditlevsen, D. E. Gardin, M. A. van Hove, G A. Somorjai,
Langmuir, 1993, 9, 1500-1503.
[67] V Bustos, M. V Gargiulo, J. L. Sales, R. O. Unac, G. Zgrablich,
Langmuir, 1997,13, 4301-4304.
[68] B. Bigot, F. Delbecq, V H. Peuch, Langmuir, 1995, 77, 3828-3844.
[69] B. Bigot, F. Delbecq, A. Milet, V H. Peuch, J. Catal, 1996, 159,
383-393.
[70] G. Blyholder, L. D. Neff, J. Catal, 1963, 2, 138-144.
[71] J. B. Butt, C. L. M. Joyal, C. E. Megiris, Chem. Ind. (Dekker), 1987,
30,3-37.
[72] C. H. Bartholomew, Appl. Catal. A, 2001, 272, 17-60.
[73] V Penchev, N. Davidova, V Kanazirev, M. Minchev, Y. Neinska,
Adv. Chem. Ser, 1973, 727, 461.
[74] J. G. McCarty, H. Wise, J. Catal, 1979, 57, 406-416.
[75] A. J. H. M. Kock, P. K. de Bokx, E. Boellaard, W. Klop, J. W. Geus,
J. Catal, 1985,9<5,468-480.
[76] F. Hochard, H. Jobic, J. Massardier, A. J. Renouprez, J. Mol. Catal,
1995,95,165-172.
[77] N. T. Dung, D. Tichit, B. H. Chiche, B. Coq, Appl. Catal, 1998,
169, 179-187.
[78] H. Knoezinger, in "Acid-Base Catalysis", VCH Verlagsge¬
meinschaft, Weinheim, 1988, pp. 147-167.
[79] M. Bodenstein, Z. Phys. Chem., 1913, 85, 329-397.
Page 134
[80] A. Chojecki, in "Selective hydrogénation of butyronitrile over
Raney-metals", Technische Universität München, 2004.
[81] O. G. Degischer, Pynitrilhydrierung Katalysatorscreening, VFCR,
F. Hoffmann-La Roche, 1999.
[82] L. J. Richter, W. Ho, J. Chem. Phys., 1985, 83, 5, 2165-2169.
[83] R. H. Newton, B. F. Dodge, J. Am. Chem. Soc, 1933, 55,
4747-4759.
[84] J. T. Yates, D. W. Goodman, J. Chem. Phys., 1980, 73, 10,
5371-5375.
[85] A. Bandara, S. Katano, J. Kubota, K. Onda, A. Wada, K. Domen,
C. Hirose, Chem. Phys. Lett, 1998, 290, 261-267.
[86] A. Bandara, S. Dobashi, J. Kubota, K. Onda, A. Wada, K. Domen,
C. Hirose, S. S. Kano, Surf Sei., 1997, 387, 312-319.
[87] A. Bandara, S.S. Kano, K. Onda, S. Katano, J. Kubota, K. Domen,
C. Hirose, A. Wada, Bull. Chem. Soc. Jpn., 2002, 75, 1125-1132.
[88] P. Claus, S. Schimpf, R. Schödel, P. Kraak, W. Mörke, D. Hönicke,
Appl. Catal. A, 1997, 7<55, 429-441.
[89] T. B. L. W. Marinelli, S. Nabuurs, V Ponec, J. Catal, 1995, 757,
431-438.
[90] M. Englisch, V S. Ranade, J. A. Lercher, Appl. Catal. A, 1997,163,
111-122.
[91] S. Galvagno, A. Donato, G. Neri, R. Pietropaolo, D. Pietropaolo, J.
Mol. Catal, 1989, 49, 223-232.
[92] C. G. Raab, J. A. Lercher, J. Mol. Catal., 1992, 75, 71-79.
[93] P. Claus, P. Kraak, R. Schödel, Stud. Surf. Sei. Catal, 1994, 108,
281-288.
[94] P. Claus, D. Hönicke, Chem. Ind. (Dekker), 1994, 62, 431-434.
[95] D. G Blackmond, A. Waghray, Chem. Ind. (Dekker), 1994, 62,
295-305.
[96] D. G Blackmond, A. Waghray, J. Phys. Chem., 1993, 97, 22,
6002-6006.
[97] E. Gindro, Dissertation Nr. 14714, ETH Zürich, 2002.
[98] H. Greenfield, D. S. Frederick, US Patent 3350450, 1967.
[99] E. Widmer, M. Casadei, Ro 86322, F. Hoffmann-La Roche, 1975.
Page 135
100] H. Ernst, Pure Appl. Chem., 2002, 74, 11, 2213-2226.
101] D. Rood, in "A Practical Guide to the Care, Maintainance and
Troubleshooting of Capillary Gas Chromatography Systems",
Wiley, New York, 1999, pp. 236-289.
102] C. J. Cowper, A. J. DeRose, in "The Analysis of Gases by
Chromatography", Pergamon Press, Oxford, 1983, pp. 225-261.
103] S. L. Morgan, S. N. Deming, J. Chromatogr, 1975, 772, 267-285.
104] "Organikum: Organisch-chemisches Gmndpraktikum", Johann
Ambrosius Barth Verlag, Leipzig, 1993, pp. 404-451.
105] J. F. Hayes, J. D. Hayler, T. C. Walsgrove, C. Wicks, J. Heterocyclic
Chem., 1996,33,209-212.
106] C. Aubert, C. Huard Penio, M. C. Lasne, J. Chem. Soc, Perkin
Trans., 1997, 7,2837-2842.
107] J. W. Wiench, L. Stefaniak, E. Grench, E. Bednarek, J. Chem. Soc,
Perkin Trans., 1999, 2, 885-889.
Page 136
Chapter 11
Appendix
11.1 Curriculum vitae
name
date of birth
citizen
Roc Novi
15.10.1977
Vignogn GR
Education
1984-1990
1990-1992
1992-1997
1997-2001
2001-2004
primary school in Savognin
secondary school in Savognin
gymnasium at the EMS in Schiers
diploma education in chemistry at the Swiss
Federal Institute of Technology in Zürich
Ph.D. thesis at the Swiss Federal Institute of
Technology in Zürich and the F. Hoffmann-La
Roche (VFCR) in Kaiseraugst
Page 137
11.2 Conference contributions
R. Novi, F. Rössler, P. Rys, "Hydrogénation of aliphatic nitriles;
thermodynamic and kinetic aspects", 6th International Symposium on
Catalysis Applied to Fine Chemicals, 6.-10. April 2003 in Delft, NL.
O. G Degischer, R. Novi, F. Rössler, P. Rys, "Application of a
hydrogénation catalyst modified by formaldehyde", 18th North
American Catalysis Society Meeting, 1.-6. June in Cancun, MEX.
Page 138